CHICAGO--"Reinventing Drug Discovery," a new study by Andersen Consulting here, notes that major pharmaceutical and biotech companies are expecting massive increases in the productivity of drug discovery activities over the next few years, and "integration of genomics-related activities into the target identification process and well-managed external alliances for genomics and target validation will be critical components of success."
Based on a survey of 100 executives and managers at 10 companies in the US and Europe, the study notes that by 2000, 60 percent of targets are expected to be derived from genomics, compared to only 8 percent in 1996, when the survey was done. Pharmaceutical industry investment in genomics had grown to over $1 billion by last year, but the investments by individual companies vary tremendously, the report said.
Another factor identified by the study as critical to companies seeking an edge in drug discovery is strategic information management, which the reported called "a powerful and untapped source of competitive advantage."
"None of the surveyed companies convincingly demonstrate a coherent strategic information technology plan that provides a fully integrated look at the architectural and applications needs for discovery information," it concluded. The study found a similar lack of a "comprehensive application architecture which can support the individual process areas while establishing the ability to share data across the organization," observing that "no company surveyed yet has a truly integrated approach across all research applications." Applications are being developed or used to support bioinformatics and other lead identification and optimization activities, it remarked, but "few companies have adequate systems to manage the inventory and flow of compounds from their libraries."
Strategic alliances will also be an important feature of the new drug discovery industry, the report said, with survey respondents expecting "to allocate 20 percent of their discovery budgets to external alliances by 2000--up from 4 percent in 1994." Although collaborations can be a highly efficient research strategy, their cost is rising, it found, and "an additional challenge repeatedly cited by research managers is in managing these collaborations effectively and balancing external collaborations with internal efforts."
The overall amount of the companies' R&D budgets devoted to drug discovery--currently an average of 30 percent--is expected to remain fairly constant in the future, even as the productivity of the efforts is supposed to rise significantly, to the point that companies expect to "require 45 percent fewer people for each new chemical entity delivered to development and a 38 percent decrease in discovery spending per new chemical entity." One critical factor will be on the personnel side. In that area, "goals, measures, and rewards that weight the innovativeness, and not just the quantity, of discovery output are effective elements," the study noted, citing one company's approach of "developing a point-scoring mechanism to reward teams for producing innovative new chemical entities.
In addition, "strong multidisciplinary teams are essential for achieving project focus, flexibility, and coordination among therapeutic areas, processes, and functions." Success seems to depend on team empowerment from a budgetary and decision-making perspective, as well as dedicated members and team-based rewards, rather than on the size or number of teams. The study found that in 1996 discovery-area headcount at the pharmaceutical companies was 380-1,300, while at the biotech participants it was 100-150.
In terms of the skills mix, only 1 percent of the biologists working in discovery were identified as molecular geneticists or bioinformaticists. The largest group were pharmacologists, biochemists, and cell biologists (75 percent), followed by molecular biologists (13 percent), microbiologists (10 percent), and biophysicists (also 1 percent).
The current ratio of chemists to biologists at the surveyed companies is 1 to 1.6, but several respondents predicted movement towards a 1 to 1 ratio in the future.
With optimization of lead identification a major goal, "traditional lead identification processes can be shortened by a half to two-thirds by developing an integrated high-throughput screening infrastructure that links compound handling, assay development, robot utilization, and informational technology systems." Because "screening library compounds must be as diverse as possible," many companies will have to buy or build "larger and more diverse libraries which will increasingly rely on combinatorial chemistry as part of the mix."
In the drive to improve productivity, "discovery organizations can reduce lead optimization time by decreasing the number and length of feedback loops between chemists and biologists and creatively organizing chemistry support to maximize the conversion of hits to leads," according to the report, which noted that at one of the companies in the survey, "improved communication mechanisms between chemists and biologists have shortened the feedback loops during analoging and secondary screening by more than half." On the flip side, "inadequate integration of late-stage discovery and early development is blamed for up to 30 percent of project time delays and up to 75 percent attrition rates at the end of Phase I," the survey found.
To realize the maximum potential of genomics in helping to deliver the next generation of drugs, "the real key lies in true functional genomics, not mere sequence data," the report found. "Within several years most sequence information is expected to be in a public gene bank, rendering proprietary sequence capabilities less valuable. Instead, companies will be making greater investments in functional genomics companies and technologies," it observed.
In conclusion, the study found that "R&D leadership must take an integrated view of discovery which embraces new thinking around discovery strategies, systematic processes, leading-edge technologies, tailored organizational approaches, and more robust information management capabilities." The result could be shortening the time required for taking a drug through discovery to the development phase from the current four to seven years to two to four years, and tripling the number of leads a typical drug company sends to development, from four to five per year per 1,000 employees today, to 14 by 2000.