Custom Chemicals

Custom synthesis providers greet new year with cautious optimism, focusing on innovation as a key component of success in a hypercompetitive business climate

The past four years have been miserable for many custom synthesis providers. The industry had expected that unprecedented productivity of pharmaceutical pipelines would lead to massive outsourcing of manufacturing, which would lead to 15% annual growth in business. Reality, however, has been far different: The pharmaceutical industry did not deliver the mid-1990s promise of 60 to 80 new products per year. In fact, demand for custom manufacturing slumped as the industry consolidated.

Since 2001, competition among custom synthesis providers has been cutthroat. For 2005, industry players and observers are cautiously optimistic, expecting another difficult year but hopeful that the long-awaited recovery really is just around the corner (see page 52). Without exception, companies are depending on innovation to survive and succeed, according to attendees of the European Fine Chemicals Conference held this past November in Newcastle, England, and the annual CPhI Worldwide exhibition, held last month in Brussels.

Pharmaceutical pipelines are encouraging, and the number of clinical projects is on the rise, according to Marc A. Hennebert, a managing partner at Strategic Decisions Group. New drug launches could boost custom manufacturing in the period 2006–08, he predicted in Newcastle. "There's light at the end of the tunnel, but we are not out of the tunnel yet," he said.

Recent painful adjustments, with many players restructuring to adjust their cost base to market realities, have changed the way custom synthesis providers view the marketplace. "From the late 1990s to the early 2000s, most companies focused on winning commercial-scale projects. The market was still growing significantly, and pharma companies were moving to outsource products and intermediates that they typically manufactured in-house," Hennebert told C&EN. Those trends fed an acquisition frenzy that saw independent players "acquired at valuations that make people laugh today," he said. For example, he pointed to the acquisition of BTP by Clariant in January 2000 for 14.9 times its earnings before interest, taxes, depreciation, and amortization (EBITDA); of ChiRex by Rhodia in July 2000 for 14.7 times EBITDA; of Catalytica by DSM in August 2000 for 11.7 times EBITDA; of Laporte by Degussa in December 2000 for 10.3 times EBITDA; and of Ascot by Dow in March 2001 for 11 times EBITDA.

Since then, changes have occurred. Competition has shifted to earlier stages of product development. Emphasis on current Good Manufacturing Practices (cGMP) and regulatory compliance has intensified, better positioning Western companies for advanced intermediates and active pharmaceutical ingredients (APIs) while leaving lower cost Asian players to compete for building blocks and early intermediates. Companies have gained a better understanding of the market, allowing them to focus on segments of their choice, and have turned low-cost Asian manufacturing from a threat to an advantage by sourcing early synthesis steps to partners, or their own capacity, in China and India, Hennebert said.

Contracts also are changing, Hennebert continued. Manufacturers increasingly are demanding agreements that require the customer to pay a fee in case of project cancellation. A recent example he pointed out is the payment of up to $27 million by Genta to Avecia in 2004. Avecia had been manufacturing the API for the oligonucleotide drug Genasense under a long-term agreement. The contract was canceled after Genasense failed to get marketing approval from the Food & Drug Administration.

Major challenges remain. Capacity still exceeds demand, making further consolidation between U.S. and European players highly likely, Hennebert said. Asian suppliers are moving into high-value-added products, which used to be the exclusive domain of Western producers. The pharmaceutical industry itself is under pressure with the withdrawal of Vioxx and the resulting cloud of uncertainty over similar drugs, such as Celebrex and Bextra.

European producers must contend with other problems (see page 58). The diminishing dollar is wreaking havoc. A discrepancy in regulatory requirements for suppliers based outside Europe is threatening the competitiveness of those manufacturing in Europe. Industry insiders are worried about a looming shortage of people skilled in science and engineering.

Companies have reorganized, restructured, cut jobs, closed plants, reconfigured business models, and even shuttered custom synthesis business units to adjust to the hostile environment. For the short to medium term, these measures could help revive the industry, according to Enrico T. Polastro, of Arthur D. Little Benelux, Brussels. In Newcastle, Polastro suggested that the long-term viability of custom manufacturing in the West would require steps that are more fundamental than mere cost cutting. "Embracing innovation in every aspect of the business is one of them," he said.

INNOVATIONS IN technology always have been central to competitiveness, but companies increasingly are taking a broader view. For example, in Newcastle, Ian Shott boldly suggested that fine chemicals companies should help the pharmaceutical industry innovate away from current wasteful drug discovery and product development processes. Shott is chairman of Excelsyn, a recently launched U.K.-based management consulting, molecular development, and engineering technology company.

Current drug discovery--based on huge libraries of molecules screened by high-throughput robots--is like "having people with Uzi machine guns mowing down people in front of them and then picking out the bodies that curiously either took a long time to die or didn't die at all. It is molecular carnage," he said.

"We need to make drug design more rational," Shott advised. "We need to think about smaller libraries of compounds with better predictive knowledge of efficacy and built-in manufacturability. We need synthesis pathways consisting of a maximum of five reaction sequences that are designed not by medicinal chemists working in isolation but by teams who understand kinetics, enthalpy, entropy, catalysis, and materials science. We need to conduct those reactions in fast reaction systems with efficient separation. We need specific syntheses at the outset so we don't spend so much time separating out what we really want."

Although pharmaceutical executives decide discovery strategies, they can be urged to rethink their approach, Shott believes. The strategy based on indiscriminate synthesis and current robotic technology has failed; in the end, market pressures on pharmaceutical companies will be the driving force, he said.

For the fine chemicals industry itself, Shott also has a prescription: improve process efficiencies through reactor design, product design, better technology for purification and separation, wider use of catalysis, cost-effective engineering, and low-cost manufacturing equipment. "The way forward does not depend on chemistry alone," he said. It depends on combining different disciplines--including chemistry, chemical engineering, control engineering, materials science, and biotechnology--and combining them earlier when there is still time to have an impact.

Shott's grand vision is already being implemented in modest incremental steps. At the Newcastle conference, improving process efficiencies was a major topic. Speakers discussed switching from batch processing to continuous processing, shrinking reactor volumes to microscale, and engaging in cross-disciplinary collaborations for process improvement.

Fine chemicals production typically occurs in a batch environment, and changing the mind-set to embrace production in a different setting is difficult. Continuous processing can reduce operating expenses by at least 90% and capital expenses by at least 50%, according to Chris Dowle, director of advanced processing at the Center for Process Innovation, in northeast England. Those data make a compelling case for a technology shift.

James Robinson Ltd., a specialty chemical manufacturer headquartered in Huddersfield, England, likely will be making such a shift. The company produces dyes and photographic chemicals. According to its managing director, Brendan Catlow, the company already had established a joint venture in India to manufacture its mature products.

SEEKING TO KEEP as much manufacturing in Europe as possible, the company turned to continuous processing using an oscillating baffled reactor. The reactor consists of a tube with orifice baffles that oscillate as the reaction mixture moves through, creating intimate mixing. Despite its simplicity, the reactor has an enormous impact on cost, time, safety, and the environment, Catlow said.

The company does not yet use the reactor for commercial-scale production. "But when we get this to work, the impact on our business will be huge," he said. "It will lead to a fundamental shift in our future asset strategy. When products approach the end of their life cycle, no longer would we inevitably have to put them out to India or withdraw them from the market. Some of them we will be able to keep here."

Meanwhile, at the CPhI Worldwide exhibition, continuous processing also was being touted. For example, Magnus Hrrd, chief executive officer of Hrrd Research AB, met there with potential clients for his company's technology: continuous catalytic hydrogenations in supercritical fluid. These are faster and more productive than conventional hydrogenations because the hydrogen, substrate, and solvent are in one phase, he said.

Originally developed to hydrogenate edible oils without producing trans fatty acids, the technology also has applications in fine chemicals production. "Whenever you have a hydrogenation and you can add a solvent, we can be there," Hrrd said. That includes reduction of carbonyls to alcohols and reduction of nitriles to amines. In these applications, the technology could deliver a three-order-of-magnitude improvement in productivity compared with batch reactions, he said. Hydrogenation of prochiral carbon centers being a key route to chiral molecules, most likely Hrrd is especially targeting chiral building blocks, but he declined to be specific. He would say only that at CPhI he met with representatives from Avecia and other companies.

Also at CPhI, SK Energy & Chemical told C&EN that it is now producing commercial quantities of (S)-3-hydroxy-

-butyrolactone by a completely continuous process from l-malic acid. The chiral compound is a starting material for a key intermediate to statin drugs such as atorvastatin (Lipitor) and rosuvastatin (Crestor). "We now can make 100 metric tons per year," said Jaeyon Yoon, director of business development for pharmaceuticals and fine chemicals.

With the success of continuous processing for this product, SK Energy & Chemical is developing other continuous catalytic reactions, including oxidations, dehydrations, and esterifications. It has a contract research arrangement with a major pharmaceutical company to look into converting current manufacturing in batch reactors into continuous processes, Yoon said. Especially for products required in hundreds of metric tons, continuous production has many advantages. The most obvious is plant size. Producing hundreds of metric tons of anything with batch reactors would require a huge plant.

In a related development, SK Energy & Chemical has begun using microstructured reactors to produce subkilogram quantities of a chiral acetyltetrahydrofuran with an eye toward commercial-scale production. The target, an intermediate to an animal-health product, is made from a chiral cyanotetrahydrofuran through a Grignard reaction. In batch mode, racemization of the center of chirality by the Grignard reagent is a serious side reaction. In a microstructured system, the reaction can be controlled precisely to avoid that side reaction. "We have shown the customer that we can make this material at a cheaper price with this technology," Yoon told C&EN.

At the Newcastle conference, too, interest in microstructured reactors was high. Henrik Hahn, head of the new project house for process intensification at Degussa, described a multiparty project to develop and build a microstructured reactor and operate it at pilot-plant scale. The model reaction for the project was heterogeneously catalyzed gas-phase epoxidation of propene with vaporized hydrogen peroxide. The reactor ran stably and reproducibly, proving that microstructured reactors can be used for commercial-scale production, he pointed out. The next step is to optimize reactor manufacturing, he added. "What we have now is handmade. What we have to come up with is commercially available reactors."

Access to expertise in microstructured reactors and other leading-edge technologies is available from the Fraunhofer Institute for Chemical Technology (ICT), in Pfinztal, Germany, said Stefan Lbbecke, vice director for energetic materials. The institute is one of 58 that make up Fraunhofer Gesellschaft, the largest organization for applied research in Europe. At ICT, "we have very different types of microstructures made with different materials and capable of different functionalities," he said, emphasizing the broad range of tools available to parties seeking assistance.

Microstructured reactors have well-known advantages, including the ability to run hazardous reactions safely, to produce reactive or hazardous species at the point of use, and to control reaction parameters precisely. With integrated monitoring, they also provide a wealth of information about the chemistry going on inside. For example, monitoring with an infrared camera reveals hot spots in exothermic reactions and also allows calculation of fluid dynamics within the channels, Lbbecke explained.

SUCH MONITORING can spring surprises. For example, in the synthesis of a dinitro-substituted urea, continuous infrared monitoring yielded not only kinetics and mechanistic information but also an intermediate that nobody expected, Lbbecke said. The new intermediate, he added, allowed a completely new route to the synthetic target.

Back at CPhI, innovation manifested itself not only in new technologies and new capabilities but also in new business models, all aimed at enhancing value for customers.

As Ralf Pfirmann, senior vice president and global business director for Clariant's pharmaceuticals business unit, put it to C&EN: "We constantly innovate customer value. Innovation can be in the business model or in understanding customer concerns. Most of the time, it's thinking ahead about what the customer might need tomorrow."

Like many others, Clariant was caught up in the frenzy of the late 1990s that brought so many new and inexperienced players into pharmaceutical custom manufacturing. It is one of several industry players that have downsized and restructured, closing plants, concentrating production in the most efficient plants, and installing high-demand technologies.

Those changes are now producing results. For example, because of high demand for its cGMP catalytic hydrogenation facility in Origgio, Italy, Clariant is expanding that capacity by 20%. The new capacity, which will be available by mid-2005, is already fully booked, Pfirmann said.

Catalytic hydrogenation is a cornerstone in the manufacture of APIs. Many synthetic routes are designed with hydrogenation at or near the end, when most of the target molecule is in place. Conditions must be mild and highly selective. "We have been mastering this technology for some time now, and the demand for it is strong," Pfirmann says.

In another area, Clariant has added the capability to manufacture controlled substances at its Springfield, Mo., facility. The first project is a central nervous system drug classified as Schedule II under the Controlled Substances Act of the Comprehensive Drug Abuse Prevention & Control Act of 1970. Clariant is preparing the API all the way from raw materials. The customer might be a generic pharmaceutical company, because, according to Pfirmann, the API "could appear in our product catalog at a certain point."

Meanwhile, Hovione has invested in another tool to control the physical properties of APIs. It has installed a cGMP spray-drying capability to help address stability issues that APIs are increasingly presenting. According to Hovione CEO Guy Villax, more and more compounds in development are fragile--sensitive to light, pH, or oxygen. With spray-drying, steps can be taken not only to produce the required solid-state form but also to improve stability, including encapsulation and incorporation of binder or excipient.<p>"We're trying to bridge the gap between the chemists and the formulators," Douglas B. Hecker, sales and marketing manager for Hovione's U.S. operations, told C&EN. "With spray-drying, we can produce API particles that are closer in physical form to that required in the finished drug product. It eliminates steps that formulators may have to take--freeze-drying, milling, and sieving--before they can work with the API."

SPRAY-DRYING has been around for many years. In the food industry, it is practiced on scales of thousands of metric tons per year. It is, however, a novelty in API production, Hecker said. And it is best suited for APIs that are problematic for formulators.

Controlled crystallization, of course, is the classic tool for designing particle properties. In 2002, Accentus introduced ultrasound-based control of crystallization, dubbed C³ sonocrystallization, at the CPhI Worldwide exhibition held in Paris. Since then, the technology has been demonstrated at various scales and for different applications. It is now being installed at commercial scale in Aughinish Alumina, a refinery in Ireland that produces alumina (Al₂O₃) from bauxite, to help remove impurities during digestion of the ore with concentrated base. Closer to fine chemicals and custom synthesis, a major pharmaceutical innovator has signed up to apply C³ sonocrystallization to the manufacture of a range of products, according to David Hipkiss, Accentus' general manager for the C³ technology business.

In Brussels, Accentus launched CrystalGEM, a predictive crystallization technology that can dramatically reduce the wet experimentation required to predict the solid-state behavior of compounds under conditions relevant to the production of APIs and other fine chemicals. The method is based on pattern recognition software and a database of crystallization systems for thousands of compounds. It was developed by George E. Tranter, director of Chiralabs, a small R&D service company with expertise in chirality and physicochemical properties. The database contains comparable crystallization information from drugs, druglike molecules, and biomolecules under various conditions, he told C&EN. The recognition software identifies patterns of certain compounds crystallizing in certain ways under certain conditions. Given a new structure, it examines how those patterns apply and predicts the crystallization behavior.

The predictions are coded to indicate which crystallization conditions must absolutely be tested in the lab and which ones are complete wastes of time. "The proof of the pudding is when we compare our prediction to what we observe," Tranter said. With indomethacin as the test compound, 86% of the conditions predicted to induce crystallization actually did produce crystals, he told C&EN. "We're getting a high rate of success by narrowing the range of conditions."

CrystalGEM can be used at any stage of drug development and product life cycle, Hipkiss noted. "If you're in discovery, you want to know how many polymorphs are viable. In chemical development, this can address issues of habit and shape for optimal bioavailability and production. For the production guys, they want short fat rods that are quick to produce and easy to filter. If you're a generic company looking for a different stable polymorph, this tool is powerful."

Accentus offers CrystalGEM as a service on the basis of an agreement with Chiralabs. The agreement is not just about Accentus being able to market CrystalGEM and Chiralabs being able to access customers it could not reach on its own. Also built into the agreement is a mechanism for synergy. The database now is structured around conventional crystallization. When information from Accentus' ultrasonic crystallization technology is incorporated, as both parties intend, CrystalGEM should be able to predict what compounds can benefit from ultrasound treatment simply from structure.

Technology-based synergistic collaboration, such as that between Accentus and Chiralabs, seems to be an increasingly preferred model for innovation. Instead of everyone inventing a different wheel, two or more companies agree to develop one that would be much better than anything one company could have done by itself or that could be brought to market faster as a joint effort.

Last year, Massachusetts Institute of Technology, Rhodia Pharma Solutions, and Bayer Chemicals AG (now Lanxess) agreed to commercialize aromatic-bond-forming chemistry invented by MIT chemistry professor Stephen L. Buchwald (C&EN, Sept. 6, 2004, page 62). At CPhI, three companies announced joint efforts to develop asymmetric processes and chiral building blocks.

CSS, a small-scale pharmaceutical contract research and synthesis organization based in Northern Ireland, has signed bilateral deals with two chiral chemistry companies. The agreements enable CSS--with its strength in synthetic route development--to offer custom synthesis of chiral compounds using either the chemical asymmetric hydrogenation technology of Chiral Quest or the biocatalytic asymmetric reduction technology of IEP. In partnering with Chiral Quest and IEP, "we've chosen people with technologies that are proven, that are complementary, and that will work for our customers today," David Moody, vice president for commercial operations at CSS, told C&EN.

IEP specializes in using alcohol dehydrogenases to make chiral alcohols. According to CEO Ortwin Ertl, IEP technology is being used in six commercial-scale projects, including two for the enantiomers of 4-chloro-3-hydroxybutyrate, one of which is an intermediate for statins, and one for (S)-2-butanol, an intermediate for a different API. "What distinguishes us is that we have extremely efficient processes. Our ballpark cost for an asymmetric reduction is $10 per kg. Given that the product is a very high value alcohol, the process is extraordinarily competitive," Ertl told C&EN at CPhI. So competitive is IEP's technology that, within a year, it turned around one company's position in the supply chain for a blockbuster API from nearly getting kicked out to being at the top of the heap.

Rtgers Organics had been making a chiral intermediate for a blockbuster drug through chemical synthesis at a price of several hundred dollars per kilogram, explained John C. Wetzel, director of fine chemicals. "Our customer had overcommitted on the supply side and was looking for reasons to throw out suppliers. We had to struggle, as the price needed to come down again and again."

IEP's technology enabled Rtgers not only to improve the material's enantiomeric purity from 98.5% to greater than 99.5%, but also "as the purity went up significantly, the processing cost came down," Wetzel said. Rtgers now leads the competition in purity and price, he claimed.

Notable, too, is the speed of technology transfer. Talks between IEP and Rtgers began in May 2001, Ertl recalled. The process, which is now being executed in Mannheim, Germany, was producing multikilogram quantities by mid-2001 and ton quantities before the end of 2001.

Meanwhile, Chiral Quest has been proving the mettle of its technology since it became a publicly traded company in 2003 (C&EN, May 5, 2003, page 54). Reliable supply of chiral catalysts is key in their industrial use. Chiral Quest has been successful in developing routes to commercial quantities of the ligands invented by the company's founder, Xumu Zhang, a chemistry professor at Pennsylvania State University.

In mid-2004, the ligand C₃ TunePhos was produced at commercial scale for Phoenix Chemicals, Ronald Brandt, Chiral Quest's CEO, told C&EN. "Phoenix made their own announcement that they were supplying a side chain to Pfizer for atorvastatin," he added. "One could connect the dots." Since then, Chiral Quest has scaled up a second catalyst, Binapine, this time for a major pharmaceutical company with a drug in clinical trials. A third ligand, TangPhos, was scaled up for a biotechnology company developing a small-molecule drug. Scale-up of a fourth ligand, DuanPhos, will be completed by the end of January, this time for a fine chemicals company working with a pharmaceutical client.

Chiral Quest, CSS, and IEP also have formed a consortium. An immediate goal is to identify and develop chiral building blocks for multicustomer use. Another is to apply new technologies to old molecules for product life-cycle management.

The consortium will identify as many as 10 multiple-use, difficult-to-make chiral building blocks that will be of use to as many as 50 customers. "We've nearly completed an exhaustive exercise in deconstructing the structures of launched and advanced development drugs and identifying common substructures," Moody explained. The next step will be product-to-technology mapping--that is, defining the intellectual property and technology that can be applied to the targets.

SUCH MAPPING is not easy, but it offers enormous opportunities in updating technology and managing product life cycle because tools are now available that did not exist 20 years ago, when many of the target structures were invented. "Once we've done the mapping, we will develop and patent these solutions, and we will do this on our own coin," Moody said.

When it comes to innovation, custom synthesis providers undoubtedly have what it takes to survive and succeed. Ultimately, however, survival and success depend on the fortunes of customers. Major drug companies continue to be the biggest customer for custom synthesis, but for how long?

Merck's withdrawal of Vioxx will have a long-term impact on custom synthesis, Clariant's Pfirmann pointed out at CPhI. "A large customer company has lost a very large cash cow. Other large customers have the same cash cows or are planning to launch similar cash cows. These cash cows finance the R&D machines. If the R&D machines have less money to spend, we will all see fewer projects."

Furthermore, criticisms of FDA in the aftermath of the Vioxx withdrawal, worries of class-action lawsuits against Merck, and emerging concerns about other pain-relieving drugs such as Pfizer's Celebrex and Bextra and the generic drug naproxen could give rise to a go-slow mind-set among drug developers and regulators. A general slowing down of drug development also will reduce the projects outsourced to custom synthesis providers, Pfirmann pointed out.

On the other hand, other trends in the pharmaceutical industry--including the low rate of new product introductions, the doubts about the continuing viability of the blockbuster strategy for big pharma, and the smaller size markets of many of the new medical advances--portend the decline of big pharma as a customer, according to Christopher Drew, manager of SORIS, a nonprofit networking service for U.K.-based chemical companies. In Newcastle, his advice to custom synthesis providers was this: Forge new relationships, find markets outside big pharma, and "think upstream and downstream of where you are" to create new opportunities.

Indeed, as Polastro pointed out in Newcastle, "the customer slate is becoming increasingly fragmented." Medium-sized companies, start-ups, generic houses, and specialty pharma companies have diluted big pharma's dominance. Several hundred customers outside of big pharma exist, Polastro said, with different business models and therefore unique custom synthesis requirements. His advice to custom synthesis providers was this: Rediscover the virtues of customer segmentation; adapt and innovate.