Outsourcing R&D in Asia: A Case Study of Pfizer

Outsourcing manufacturing and clinical-trial materials is an important part of pharmaceutical companies' strategy for drug development and commercialization. The outsourcing model, however, may be applied earlier in the drug-development continuum to include research and development (R&D). Gonghua Pan, associate director and head of the parallel medicinal chemistry (PMC) sourcing operations at Pfizer (New York), explains the company's approach to outsourcing R&D, including the role of offshore contract services providers.

STOCKBYTE

Drivers for increased outsourcing

On an industry-level, Pan outlines the drivers for increasing R&D outsourcing. These include:

• Access to additional R&D capacity

• Access to expertise

• Access to novel enabling technologies

• To focus internal resources on core competencies

• To accelerate speed of development

• To obtain operational flexibility

• To reduce fixed costs.

Origins of research sourcing at Pfizer

Pfizer first began outsourcing certain R&D functions in the late 1990s. At that time, Pfizer had a strategic goal to increase the size of its compound collection generated by PMC techniques. PMC refers to using combinatorial chemistry and high-throughput screening to generate large libraries of compounds for testing for biological activity against a disease screen.

"PMC technology at this time was viewed as a specialized technology that was platform-centric, capital-intensive, and IT-intensive, which required a specialized skill set," explains Pan. "The company used strategic alliances to gain access to this technology."

In 1999, Pfizer signed multiyear collaborations to access PMC technology with four service providers: the UK business of Tripos (St. Louis, MO), Discovery Partners International (DPI, San Diego, CA), Arqule (Woburn, MA), and ChemBridge (San Diego, CA). Pfizer ended its drug-library agreements with three firms (Arqule, DPI, and Tripos) in 2005.

During the early 2000s, the business environment for contract drug-discovery services evolved, and several of Pfizer's early collaborators exited the market. Arqule, one of the largest contract research organizations (CROs) in the US, exited the contract-chemistry services business in 2005 to focus on developing its own oncology drug portfolio. DPI sold its drug-discovery service operations to Galapagos (Mechelen, Belgium) in 2006. Tripos sold its drug-discovery informatics division to Vector Capital in 2007, which created a separate and new privately held company Tripos Discovery Informatics (St. Louis, MO).

Transactional chemistry in Asia

As the US marketplace changed, "there was a desire to redirect our PMC resources to complement our efforts in Asia," says Pan. Pfizer established an R&D center in Shanghai in 2005. In assessing the potential of CROs in India and China, Pan explains that certain services are more suited to outsourcing when working with CROs in Asia. "Although some activities lend themselves well to sourcing in Asia, they tend to be activities that tolerate longer lead times," says Pan. These activities include developing drug libraries for use in a corporate collection, creating custom monomers for future use, and producing and resupplying bulk intermediates.

As the process moves forward to include derivatization of lead compounds and further biological testing, cycle times increase and the distance and time involved in performing these functions become a consideration.

"In-line analog production requires cycles of synthetic chemistry and in vitro testing to advance matter to development," explains Pan. "Early stages such as 'explore, screen, and design' and 'screening, design, and synthesis' are more cycle-time forgiving." In this part of the drug-development continuum, more analog series are pursued in parallel. There is less information on which leads and analog series are preferred, and matter can be assessed (either determined to be valued or terminated) with two to five assays, explains Pan. "Later stages (e.g., lead development and candidate selection) are less suited to remote analog production," says Pan. At this stage, programs often are narrowed to one to two analog series. Additional and more specialized assays are used. And process development and fine tuning of the route are common, explains Pan.

"Initially, time was a big downside to sourcing R&D in Asia," says Pan. The company had to consider the time to ship material to a CRO's site; the time to move product from China to Pfizer's research sites in the United States and the United Kingdom; the time to receive the material back at the CRO's site; the time to process the material; the time to test; and the time to upload data. The emergence of screening capabilities in Asia, however, changed the outsourcing paradigm for contract R&D. "The prospects of localized screening allowed for a virtual pharma model by which Pfizer could design the drug-screening criteria, but the execution of this criteria could be made in Asia. The ultimate outcome of this process was validated lead matter," says Pan.

From line-centric to integrated sourcing

As the capabilities of its CROs improved, Pfizer evaluated three potential operating models for outsourcing R&D as outlined below:

• An integrated model under which all activities (e.g., discovery chemistry, assays, screening, in vitro testing, and absorption, distribution, metabolism, and excretion [ADME] testing) are consolidated at a single CRO

• A rationalized model under which multiple CROs provide chemistry and biology services

• A diffused model under which three different CROs provide chemistry functions (e.g., discovery chemistry), biology (e.g., assays and screening), and ADME testing (e.g., ADME and in vitro testing).

Pan outlines the advantages and disadvantages of each approach. The integrated model offers logistical advantages, lower cycle times, and the ability to develop a relationship with an established CRO. The disadvantage of using an integrated model is possible leakage of intellectual property (IP) because more functions are consolidated with a single CRO. "There was organizational resistance to this model because we historically divided functions among CROs as a way to prevent IP leakage. We had concerns of limiting our choice to essentially one CRO," says Pan.

The rationalized model offers the advantage of enhancing Pfizer's ability to manage IP. By working with more than one CRO, this approach distributes the risk and is more competitive, says Pan. The disadvantages are that the rationalized model increases the complexity of the project because it involves working with more than one CRO. This approach also lacks a single point of accountability for logistical issues among CROs. Also, this approach would require upfront investment in screening and ADME capabilities by Pfizer, if pursued.

The last approach, the diffused model, requires the least investment to develop capabilities and provides Pfizer the ability to choose the "best-in-class" CRO for each task, explains Pan. The drawbacks, however, are higher complexity and longer cycle times resulting from working with multiple CROs. "This approach would also require increased internal management and decrease our ability to develop or shape a CRO," says Pan.

Building the partnerships

After evaluating these options, Pfizer decided to shift its R&D outsourcing strategy from a "line-centric model to an integrated sourcing model," explains Pan. The integrated model favors working with one CRO that is able to provide chemistry, biology, and ADME screening and testing services. "Companies such as WuXi PharmaTech were identified as prototypical partners for such complicated services," says Pan.

Pfizer partnered with WuXi PharmaTech, a Shanghai-based CRO, in 2006. Pfizer and Merck & Co. (Whitehouse Station, NJ) were WuXi PharmaTech's two largest customers in 2006, respectively, accounting for 15.4% and 13.7% of the company's 2006 net revenues of $69.9 million. Pfizer, Merck, and Vertex Pharmaceuticals (Cambridge, MA) were the company's three largest customers in 2007, respectively accounting for 15.0%, 12.2%, and 11.2% of the company's 2007 net revenues of $135.2 million (1).

WuXi PharmaTech's operations are grouped into two segments: laboratory services, consisting of discovery chemistry, biology, toxicology, pharmaceutical development, analytical services, biopharmaceutical and medical-device testing, and other related contract R&D services; and manufacturing services, focusing on advanced intermediates, active pharmaceutical ingredients, and biologics-based manufacturing, testing, and related services. The company's discovery-chemistry services include lead generation, which involves designing and synthesizing libraries and templates for library synthesis, benchmark-compound synthesis, and custom synthesis. The company uses a traditional medicinal chemistry approach, a process through which a series of compounds are designed with computational chemistry and related structure-activity relationships, information analysis, followed by their chemical synthesis, biological activity and ADME property evaluation. The company's synthetic-chemistry services involve synthesis of assay standards and benchmark compounds (1).

The company's primary China-based facilities include a 630,000-ft² R&D center in Shanghai, a 220,000-ft² process-development and CGMP manufacturing plant in the Jinshan area of Shanghai, and a 130,000 ft² R&D center in Tianjin, which is mainly focused on discovery-chemistry services. Earlier this year, WuXi PharmaTech acquired AppTec (St. Paul, MN), a provider of analytical testing, contract R&D, and biologics-manufacturing services. The acquisition gave WuXi three US-based facilities: a 63,000-ft² R&D and manufacturing facility in St. Paul, Minnesota, a 46,000-ft² testing facility in Atlanta, Georgia, and a 75,000-ft² R&D, testing, and manufacturing facility in Philadelphia, Pennsylvania. The company is building a preclinical drug-safety evaluation center in Suzhou, which it expects to be operational in 2009. WuXi also recently completed a 22,000-ft² CGMP pilot laboratory, located near its main facility in Shanghai. The facility is focused on formulation projects for Phase I and II clinical-trials-material manufacturing. In January 2007, the company began an expansion of its Jinshan facility by adding approximately 350,000 ft², which will quadruple the capacity of its CGMP-quality facility with an additional 172,000 L of reactor volume. The new facility will concentrate on commercial production with advanced automation and cryogenic capability. The expansion is scheduled to begin operations in late 2008 (1).

In addition to WuXi PharmaTech, Pfizer has developed relationships with other CROs in Asia. These CROs include Chembiotek (Kolkata, India) and HD Biosciences (Shanghai). Chembiotek provides biology, chemistry, and preclinical research services at multiple sites in India, and HD Biosciences provides drug-discovery contract services.

In working with these CROs, execution is key. "We invest in local logistics and informatics to enable our pilot programs with these and other companies," says Pan. Drug samples are designed by Pfizer. Its chemistry partners in Asia prepare the samples and process them locally in Asia to enable downstream screening. This sample preparation and processing involves preparing plates for in vitro biology and ADME screening. The mother plate is kept locally in Asia in case additional material is needed. The remaining material is shipped to the US for future use.

Lessons learned

Pan highlights several positive developments from these initial pilots. "After high initial start-up investments in time and resources, the effort for subsequent programs goes down rapidly. We also learned that generalized ADME assays are a value multiplier for chemistry functions. Also, we learned that implementing third-party–third-party collaborations is possible."

Despite these positive signs, Pan says certain issues need further consideration. "There were high start-up costs associated with data and logistic integration. We find that expertise in data and logistic integration lags scientific on-boarding. We also find it difficult to find local laboratory talent with a track record and the skills needed by Big Pharma." In addition, the supply chain for reagents is not as advanced in supporting broad biological screening and testing needs.

Pan says Pfizer will continue to evaluate strategic sourcing pilots for R&D in Asia that go beyond traditional chemistry-related services to include activities in expression and purification, structural biology, screening, bioanalytical testing, ADME testing, toxicology, and molecular profiling. These pilots will explore R&D capabilities in Asia by evaluating the following:

• CROs in India and China

• Programs requiring good laboratory practices (GLPs) and non-GLP programs

• Functional-driven pilots and integrated pilots

• Comparisons of single-vendor pilots and multiple-vendor pilots

• Strategic sourcing in multiple therapeutic areas and sites.

Overall, Pan says the company's experience in outsourcing R&D in Asia has been positive. "We have seen explosive growth in capabilities and capacity during the past four years. Going forward, we hope to continue to push the boundaries of activities associated with strategic sourcing to help us deliver the next wave of medicines."

Reference

1. US Securities and Exchange Commission (SEC), "Form 20-F: WuXi PharmaTech (Cayman) Inc.," (Washington, DC, 2007).