The pharmaceutical industry must cut the time its drugs linger in development and terminate less promising projects earlier if it is to improve the efficiency and productivity of its R&D programmes.
These are among the main conclusions of the annual Outlook report, issued by the Tufts Centre for the Study of Drug Development, which warns that "business as usual is no longer an option when it comes to developing new prescription medicines."
The decline in R&D productivity in the industry - measured by the falling number of new drugs launched on to the market in recent years despite an ever-growing spend on R&D - has been causing concern for some years.
Tufts believes there are a number of reasons for the fall-off, including a greater emphasis on developing drugs for chronic and complex diseases that require longer, larger and more costly clinical trials.
Meanwhile, it is becoming harder to recruit and retain people in studies, and regulators are demanding more extensive data on safety following a number of high profile product withdrawals in recent years.
One of disappointing conclusions of the report is that the industry's investment in often expensive drug discovery technologies such as high-throughput screening has yielded negligible returns in terms of drugs going through clinical trials and onto the market.
Tufts - which is perhaps best known for its widely reported estimate that each drug that reaches the market costs $802 million to develop - believes that drug developers will increasingly pursue joint ventures and partnering opportunities, while late-stage licensing will decline.
Meanwhile, downstream research activities such as comparative trials between drugs will be constrained due to conflict of interest concerns and restrictions on access to patient records.
R&D projects should receive greater scrutiny early in development and companies should improve their ability to spot the poor prospects using risk assessment and management methods.
Although development times for new drugs have been on the wane since the 1980s - when it typically took more than 8 years to take a drug through the clinic and to approval - additional reductions are required.
Average new drug development times are between 6 and 7 years at present, but the benefits from cutting them further are apparent. For example, Tufts notes that 48 per cent of the total clinical cost of around $530 million to develop a drug to treat central nervous system drugs relates to time.
"Given rising clinical study and related out-of-pocket costs, cutting development time offers the most potent tool for containing R&D expenditures," it said.
Aside from the need to revamp the R&D function in pharmaceutical companies, Tufts has also identified a number of regulatory trends that will affect the industry in 2004.
For example, the European Medicines Evaluation Agency (EMEA) will push forward a number of regulatory reforms, predicts Tufts. These could include the provision of more scientific advice to companies, instituting a US style 'fast-track' system to speed up the approval of new drugs for unmet medical needs, and allowing provisional marketing authorisations for experimental drugs on compassionate grounds.
And in the US, the Food and Drug Administration (FDA) will expand its fast-track system to include additional diseases and develop regulatory frameworks for emerging technologies such as gene therapy and pharmacogenetics.
Tufts' Outlook 2004 report will be placed on the Centre's website , for download free of charge, in March.