Process Development: Think Like a Scientist—Behave Like a Business

Understanding the end-to-end management of chemistry, manufacturing, and controls (CMC) resources provides the opportunity to enhance long-term planning, leverage development options, manage resource trade offs, and track progress against plans. The goal is to improve the pharmaceutical development process to deliver the pipeline. This article provides an overview of the organizational structure of Process Research and Development (PR&D) and the CMC teams at Genentech; the alignment of resources based on CMC contracts, process development activity maps and project resource plans; and the business economic analysis for evaluating development options.

Sue E. Steven, PhD, MBA

BUSINESS IMPERATIVES

Genentech, Inc., like many biotechnology and pharmaceutical companies helps patients by developing medicines through great science. The Process Research and Development department at Genentech has three key business imperatives that must be met to support the company's goals. These include (1) deliver the pipeline, (2) be efficient, and (3) innovate.

Deliver the Pipeline

Genentech makes significant investments in research and development in an ongoing effort to discover and develop novel medicines. Consequently, the pipeline at the company is growing rapidly. In addition, Genentech's product platform is expanding from one primarily based on monoclonal antibodies to one that also includes small molecules, antibody-drug conjugates, novel delivery systems and devices, and other protein products. So, from the PR&D point of view, the pipeline is really a collection of distinct "mini-pipelines." Within Genentech, the process development (PD) organization strives to keep CMC activities, where practical, off the critical path in the development of new molecules.

PR&D staff at Genentech review updated PD activity maps.

Be Efficient

The lead time to develop PR&D infrastructure (buildings, staff, and equipment) is often greater than the speed at which new molecules can exit research or be licensed in through business development. Readiness to handle the influx from the pipeline cannot depend on a scale up of staff or laboratory space. Moreover, Genentech, like all companies, strives to keep the cost of drug development in check. Therefore, efficiency is essential.

Innovate

The best way to deliver the pipeline efficiently is to innovate. Innovation may involve expansion of the PD knowledge base with cutting-edge science and technology; the use of science to influence regulatory policy; or application of the latest business practices to streamline activities and decision-making.

This article describes some of the approaches that the PR&D department at Genentech has taken while pursuing these objectives.

FUNCTIONAL ORGANIZATION

PR&D has made several fundamental changes to its functional organization.

1 Bioprocess development (e.g., cell culture and purification) was split into two groups: early-stage bioprocess development and late-stage bioprocess development. Early-stage bioprocess development focuses on developing the upstream and downstream processes needed to produce materials for Phase 1 and 2 clinical trials, and supporting studies. Today, the early-stage group is on the critical path with activities involved in the production of stable cell bank through delivery of good laboratory practices (GLP) supplies for toxicology studies. Therefore, its business focus and improvement efforts are concentrated on speed. Late-stage bioprocess development focuses on delivering the processes and materials for Phase 3 trials. This group must ensure regulatory success in a business environment in which product quality and manufacturing robustness, and efficiency are paramount. These two drivers—speed and regulatory success—result in different, but appropriate, balances of scientific and business approaches in the two groups.

Quick Recap

2 A new department of global manufacturing science and technology (GMSAT) was created. Like some other growing biopharmaceutical companies, Genentech finds itself manufacturing the same product at multiple sites in the internal and external network. The role of GMSAT was created to ensure that processes, problem resolution, and improvements are treated in a synchronous manner across sites. It mitigates risk and ensures long-term success in delivering products to patients, despite the growing complexity of the pipeline. Moreover, the new GMSAT group both drives and benefits from Genentech's Product Operations initiative to become a Class A organization.¹ (Class A is a standard of world-class performance, independent of industry class.)

3 R&D expanded to Oceanside, California. This opportunistic expansion was made possible by the purchase of two manufacturing sites (a bulk manufacturing facility and a clinical operations facility) from Biogen Idec. The opportunity the expansion presents to PR&D is two-fold. First, the intent was to have the Oceanside PR&D department organized by cross-functional teams focused on early stage development to handle the growing pipeline. This provides an immediate increase in the capacity for process development. Second, the Oceanside site serves as an "incubator" for new ideas that can be tested with a team that is small, but still has all the elements of a complete PR&D department. This department has been piloting novel technologies, innovative information-technology systems, and new equipment.

4 The role of the process economics group was expanded. Even if there is an opportunity to make improvements in a process, it does not necessarily follow that those improvements should be implemented into manufacturing. It is critical to identify the opportunities and solutions that will have the greatest impact on the business. The process economics group easily pays for itself in multiples. Some of the ways in which this group supports decision-making in PR&D at Genentech include the following:

• Evaluating crystallization versus chromatography for a new product

• Maximizing the value of developing a version 1.x of an existing product

• Analyzing the value of developing an alternative device to eliminate vial overfill

• Choosing among dozens of new technologies for long-range impact.

One of the underlying benefits accrued through a centralized business-modeling group within PR&D is the assurance of common assumptions—"apples to apples" comparisons—across projects.

Figure 1

CMC REVIEW COMMITTEE AND CMC CONTRACTS

When organizations strive for efficiencies in the drug development process, they sometimes overlook the time and effort that executives spend learning and making decisions. As a pipeline grows, the amount of time spent by senior executives per molecule must decrease. Through the judicious use of functional-area committees and agenda-driven membership review, we were able to reduce by half the number of executives required to serve on the CMC Review Committee (Figure 1). These efficiencies were made possible by clearly defining roles and responsibilities (a component of Class A methodology) and by following Genentech's standard decision-making model. This model identifies a single point of accountability—who is the decision maker for any given decision? This individual is accountable for making the decision and for seeking the appropriate, informed input from others. The discussions leading up to a decision can consist of vigorous debate, but once the decision is made, all stakeholders are responsible for supporting and implementing the decision.

Figure 2

The management and oversight of the growing number of CMC project subteams has improved by employing standardized contracts between the CMC review committee and CMC project teams, as well as through templates for presentations, benchmarks for resources and timelines, and more experienced mentors or sponsors for CMC teams. Various contracts cover different predefined periods of development. The review committee business process has been particularly effective in identifying the decision points (of a total of four from inception to launch) at which a review is needed, boundary ranges for budget and timelines, and the information and deliverables required to move to each successive contract stage (Figure 2).

Figure 3

PROCESS DEVELOPMENT ACTIVITY MAPS

PD activity maps provide a framework for describing the development progression of molecules from inception to launch. These maps delineate each functional area's discrete activities, the average amount of time it takes to complete each activity, and the typical resources needed to complete the activity (in our case, at the functional-area level, Figure 3). The initial map was based on Genentech's experience over a number of years in developing MAbs. Additional maps have been crafted for other proteins, small molecules, device development, postmarket changes, technology transfers, and antibody-drug conjugates. The maps are aligned with development phases and CMC contract stages and also with non-CMC research and development activities. These have been used to identify ways to attain greatest efficiencies. The maps are updated to reflect efficiency gains during the course of the year and are leveraged to compare year-to-year changes and their impact on long-term plans (Figure 4).

Figure 4

Some examples of improvements that have affected the resources or timelines for the PD activity maps include the following:

• Through the use of innovative technologies, the number of resources required for viral validation studies has remained relatively flat, even as the volume of studies has almost doubled.

• By defining business processes in partnership with regulatory affairs and developing submission templates, our CMC editors have nearly tripled their productivity.

• As a result of several improvements—the application of platform technologies, an increase in pilot-plant capacity, and advances in technology—the number of transient cell lines produced for research and the number of stable cell lines have significantly increased per full-time employee.

• By changing the stringency of methotrexate treatment earlier in cell-line development and developing serum-free cell lines, several months have been trimmed off the cell-banking critical path.

At Genentech, PD activity maps have become the foundation of many useful applications and processes. The maps allow resource forecasting at the project and portfolio levels, thereby facilitating both short-term and long-term planning. Thus, metrics (planned versus actual resource utilization) can be tracked, transparency and alignment between teams and functions in budgeting ensured, process improvements quantified, and project contracts simplified. In addition, these maps can be used to propose alternatives—"what if" scenarios—for drug development before a substantial investment is made. For example, they can profile a scenario in which resources are gated until positive Phase 3 data are available (Figure 5).

Figure 5

PROJECT RESOURCE PLANS

Leveraging PD activity maps, CMC project leaders have assumed accountability for leading project resource planning and reporting, as well as for staying within project budgets, as defined by their CMC contracts. Functional-area managers are still accountable for staying within the cost center budgets that they manage. To support their efforts in this regard, modular project resource plans (PRPs) have been developed as a focal tool for cross-functional alignment with functional-area managers and resource-capacity planning systems. Since PRPs are prepopulated with data from PD activity maps, the need for customization is minimized. The PRPs also facilitate agreement between the CMC project team and the functional-area managers who own the specific resources. Use of the modular PRP process within each CMC contract stage has allowed simplicity when scaling of our planning capability, and enhanced planning accuracy.

Finance Department Perspective on Changes in Process Research and Development

SUMMARY

With the implementation and integration of these new business processes, organizational changes, and decision-making frameworks, Genentech has positioned PR&D for success in delivering the expanding pipeline. PD activity maps for unnamed new molecular entities (NMEs) and PRPs for existing planned NMEs can be consolidated and extrapolated to create resource-capacity models that enhance functional-area managers' ability to anticipate resource shortfalls or surpluses, move resources across functions, or justify additional hires. These spreadsheet tools and processes prepare the organization for future enterprise-wide project and resource-management systems. The organization can then pursue the ultimate goal of portfolio resource management: end-to-end balancing of resources across new and existing products to manage risk and improve returns.

In conclusion, our scientists are developing into leaders who are capable of determining the optimum balance between science and business. The perspective of the scientist offers an expansive vision of possibilities; the business perspective helps drive the selection of choices that are profitable—ultimately benefiting our patients.

ACKNOWLEDGMENTS

Maurits Trouerbach, Ken Achacoso, Susan Smigelski, Connie Veilleux, Ka Kam, Lisa Wyatt, Qi Chen, John Joly, Amy Shen, Linda Khym, Kyle Derstine, David Chang, and Ann Lee.

Sue E. Steven, PhD, MBA, is a senior director of process development–business excellence, strategy, and training at Genentech, Inc., South San Francisco, CA, 650.4670.5439, steven.sue@gene.com.

REFERENCES

1. Oliver Wight International, Inc. The Oliver Wight ABCD checklist for operational excellence. 5th ed. New York, NY: John Wiley & Sons, Inc; 2000.