R&D COST PER AVERAGE NEW DRUG IS $200 MIL., PRODUCT LIFE IS 10 YEARS, TUFTS STUDY SAYS; FDAer SAY DON'T UNDERESTIMATE APPROVAL PROCESS CHANGES

"The mean pre-tax R&D cost for a new drug introduction, measured in '87 dollars, exceeds $200 mil," according to preliminary results of an updated version on the cost of research by Tufts University's Center For the Study of Drug Development. Henry Grabowski, PhD, of Duke University, a principal investigator for the study, told a Dec. 20 seminar sponsored by the Institute of Medicine in Washington, D.C. that the findings indicate "R&D costs are clearly increasing at a very rapid rate, and are much higher in real terms than previous estimates have indicated," even after adjustment for inflation. The new Tufts estimate, when finished, will be the most recent addition to the long-running research lotto game trying to quantify the average test tube-to-market cost of a successful drug product. The Pharmmaceutical Manufacturers Association has been citing a $125 mil. per entity figure for the last several years, based on an extrapolation from a study in the 1970's. The new conclusions are being developed, Grabowski said, from "microeconomic analysis...based on a large representative sample of drugs at different stages of the R&D process...incorporating current information about research and development times and success rates." In addition to Grabowski, the principal investigators in the study include two of the established figures in the R&D cost debate: the director of the Tufts center, Louis Lasagna, MD; and the dean of the William Simon Graduate School of Business at the University of Rochester, Ronald Hansen, PhD. It was Hansen's previous study that was the foundation for PMA's recent cost estimates. The key factor in the rise in real R&D costs, according to Grabowski, is the trend toward "a more knowledge-driven, discovery-by-design approach toward R&D, requiring longer times to commercial payoff and much larger upfront investments." He also contended that the costs of biotechnology development are relatively high, as are the costs of developing products for chronic diseases, which "require more regulatory testing, and also traditionally have had lower success rates." The higher threshold costs of developing a drug are coinciding with a decline in the average length in the commercial life of new chemical entities, Grabowski indicated. In the 1970's, the Duke prof said, a typical innovator firm maintained an 80-90% market share of the compound's total sales several years after patent expiration. When he recently examined the experience of 18 major drug products for which patents expired after Waxman-Hatch, Grabowski said he found that one year after patent expiration the average loss of market share for an innovator product was 35%, with 17 generic competitors entering the scene. In two years, the innovator had lost an average of 50% of the market. Grabowski estimated that "the sharp erosion in the post-patent period is likely to intensify in the 1990's with the cost-containment pressures presently there in the health care sector." Grabowski also estimated the time from lab studies to marketing approval in the U.S. to be about a dozen years -- a figure about two years longer than recent PMA estimates. Grabowski's calculation assumes a longer period for laboratory and animal studies: three years versus the one to two years assumed by PMA. "By 1984," he said, "the IND or clinical investigation phase averaged over six years, and the NDA phase...was about two and a half years...When one adds a preclinical phase of at least three years, one has a mean total R&D time of over 12 years by the mid-1980's," a figure that is still "in an upward trend." The patent restoration provisions of the 1984 Waxman-Hatch Act fail to remedy the situation, according to another recent Center for the Study of Drug Development cited by Grabowski. That analysis found that the average benefit in increased patent life under the transition benefit is approximately one and a half years and under the full benefit will rise to an average of three years. According to Grabowski, this means "we've been settling down into a situation where the average drug will probably have 10-12 years of product life." "As we go into the 1990's," he concluded, "the period of patent exclusivity is basically when the firm has got to recoup and earn positive returns on R&D investment...Increased real prices have been one of the main strategic responses in the pharmaceutical industry to higher R&D costs, shorter product life-cycles, and increased generic competition...these higher real prices in the 1980's played an important role in the industry's return on R&D. Citing his study's estimate that among U.S. prescription pharmaceuticals approved in the 1970's, "the mean compound earned an average real return of 9%" over its market life, Grabowski defended the 1980's faster-than-inflation increases in drug prices, stating that "if no real price increases had occured in the 80's...if drug prices had merely increased at the general rate of inflation...the present value of after-tax cash flows for the mean compound would have been reduced by 16%." * FDA's Paul Coppinger, deputy to the Associate Commissioner for Planning & Evaluation, disputed some of Grabowski's cost estimates at the IoM conference. Coppinger pointed out that "nearly 40% of the new compounds that started clinical research in the U.S. in the 1980's were discovered and survived animal testing abroad before U.S. trials. Since the dry holes and discovery costs of these NCEs rest presumably on some foreign balance sheets, I think it's important to include these drugs in the U.S. calculation of return on investment." In order to arrive at more accurate assessments of drug R&D costs, Coppinger continued, "we need to consider also the construction of more global assessments of return on investments in the pharmaceutical industry: recently, I examined the research histories of 30 NME's approved in this decade for life-threatening and seriously debilitating illnesses -- about an eighth of all FDA approvals for NME's in the 1980's. More than half of these therapies originated in foreign firms, and exactly half were approved on the basis of clinical trials conducted abroad." Coppinger also suggested that the drug development equation should include those products in-licensed and developed from a later stage. U.S. drug makers should include their returns on drugs whose rights they license or purchase from abroad for a more accurate comparison of earnings and R&D spending, Coppinger contended. In addition, the FDAer noted, 11 out of the 30 FDA-approved life-saving therapies began their clinical lives as part of government or university-funded research. Drug industry estimates of R&D costs "still don't adequately take into account publicly-funded research" particularly within therapeutic categories on which NIH efforts are focussed, he said. The FDAer also suggested that agency policy changes during this decade towards drugs for life-threatening diseases should not be "underestimated." In the case of the 30 approvals for life-saving new drugs, "the majority obtained approval with median pivotal studies that contained less than 250 patients...only 11 of these were orphan therapies. Among the 13 life-saving therapies approved under Frank Young's stewardship of FDA, NDA's were submitted an average of more than six months before completion of pivotal studies...a decisive change in the earlier industry practice of submitting NDA's an average of a year after completion of pivotals." Moreover, the approval process for these products was significantly accelerated, he said. The improved regulatory process for such breakthrough products, Coppinger stated, "has developed and approved these life-saving therapies in an average of less than five years, and has been made possible by a substantial increase in FDA's consultative involvement in the clinical development process." While admitting that the relative straightforwardness of demonstrating improved survival rates sometimes simplified the research profile for life-saving therapies, Coppinger said that examination of this category of products highlighted his point that "the one-size-fits-all calculations of average R&D costs is very misleading to the public policy purposes for which those estimates are often used." He said his "hope" was "that research, such as Henry Grabowski's, moves us quickly beyond the rather sterile and deceptive statement that 'most new drug therapies will never recover the average R&D cost.' This statement, whatever dollar figure is attached to it, is a statement more about our ignorance than our knowledge of the diverse and complex drug development process."