HAI Seminar: World Trade Organization/GATT, Pharmaceutical Policies and Essential Drugs
October 4, 1996
Center for Study of Responsive Law
I would like to take this opportunity to explain some of the activities in the United States which concern international policies toward the development of new health care technologies, and to suggest a very particular public policy approach with respect to the promotion of research and development.
I will begin with an introduction. My name is James Love. I am an economist. Since 1990 I have worked for the Center for Study of Responsive Law (CSRL) in Washington, DC. The CSRL was created by Ralph Nader in 1968 as a public interest research organization. In 1991, Mr. Nader asked me to review a contract between Bristol-Myers Squibb (BMS) and the National Cancer Institute (NCI), regarding the development of Taxol, an important cancer drug. The United States government had funded all of the important research on Taxol, and had sought a firm to commercialize the invention. After a “competitive” solicitation, BMS was chosen by NCI to bring the drug to market.
With Taxol, as with many other cancer drugs, the federal government had been responsible for bearing the costs and risks of the discovery and development of the drug. The government had used a private contractor, Hauser Chemical, to manufacture the drug. BMS simply hired a key member of the NCI research team, an NCI employee named Dr. Robert Wittes, and Hauser, the NCI contractor, and agreed to provide the NCI less than $5 million of Taxol for use in the final stages of testing on the drug, which was done in a number of NCI sponsored clinical trials. There was very little risk for BMS, because NCI had already conducted very successful human use clinical trials on Taxol.
After public outcry in the United States over the pricing of the drug AZT, the Bush Administration had adopted a policy of requiring some firms to sign “reasonable pricing” clauses in return for exclusive contracts for federal government research on pharmaceuticals. Taxol was covered by such an agreement, and when the drug was approved for marketing in the United States, I was asked by a member of our Congress to evaluate the price that BMS had negotiated with the NCI.
We found that BMS had entered into a contract with Hauser (the former NCI contractor), to produce 400 kilos of Taxol for approximately $.25 per milligram. BMS was reselling the drug to cancer patients for $4.87 per milligram, more than 19 times the company’s costs. Put another way, BMS was paying $100 million to purchase Taxol that it would sell for $1.95 billion. I did not think the prices that BMS charged cancer patients for Taxol were reasonable, given the company’s negligible role in the discovery and development of the drug.
Our work on Taxol led to a much broader investigation to the pricing of drugs developed with federal government support. We found that the U.S. government had funded clinical trials and other research for 34 of the 37 cancer drugs approved for marking in the U.S. between 1955 and 1992. We found that half of all FDA “priority drugs” approved for marketing between 1987 and 1991 had benefited from a significant federal government role in funding research on the drug.(2) We found that the government focuses its research investments on drugs which represent largest gains in therapeutic value and which treat the most severe illnesses. The private sector often develops “me too” drugs that target the most profitable markets.
We also investigated the U.S. Orphan Drug Act, which was passed to support the development of uneconomic drugs for rare diseases. The law had been amended and transformed into a new intellectual property right that is routinely used for highly profitable drugs, including drugs which serve large client populations.(3)
In the course of these investigations, we were asked to testify before the United States Congress on several occasions. Some members of the U.S. Congress were interested in the ways the government could protect consumers from excessive prices on drugs, including drugs protected by the U.S. Orphan Drug Act, or drugs developed with public support. At one such hearing, in February, 1993, before the United States Senate Special Committee on the Aging, Ralph Nader and myself were asked if the government should use compulsory licensing as a tool to protect consumers.(4) Mr. Nader told the Committee that the U.S. government’s ability to do so would be severely constrained by the pending North American Free Trade Agreement (NAFTA). This was the first time I understood that international trade agreements would limit our national sovereignty with respect to policies on pharmaceutical drugs.
In 1994 I was invited to attend a meetings on trade policies and pharmaceutical drug development in Buenos Aires and San Carolos de Barioche, Argentina.(5) Here I was surprised to learn first hand the extent to which the United States government was pressuring countries to rewrite national legislation on pharmaceutical patents. At the time, Argentina did not recognize patents on pharmaceutical drugs, and it had a large domestic pharmaceutical drug industry. The Argentine Congress was considering legislation to create a system of patents for pharmaceutical drugs, in order to conform to the TRIPS section of the new GATT. The United States government wanted terms significantly different than those the Argentine Congress was willing to approve, and the President of Argentina was threatening to impose a new patent law by executive fiat. The U.S. was asking for terms in the legislation which exceeded the country’s responsibilities under the new GATT. I was asked by members of the Argentine Congress about the authority exercised by the U.S. Presidents under Executive Orders. There was a great deal of concern over this issue in Argentina, because the government had only recently become a democracy, and the U.S. was pressuring the President to exceed his constitutional powers.
Later that year I had the opportunity to visit São Paulo, Brazil, where a similar dispute over pharmaceutical patents was unfolding. In São Paulo, the multinational pharmaceutical industry trade associations were simultaneously arguing that research and development were too costly and complex for the Brazilian government, and that the industry would bring many new research and development facilities to Brazil, if the country would strengthen its domestic pharmaceutical patent laws. I was appalled by the misinformation about drug development costs, much of it taken out of context from footnotes in U.S. government reports on the topic. At one point the Brazilian representative for the multinational pharmaceutical firms told the audience that the U.S. government said it cost $359 million to develop a new drug, and each drug was a 5,000 to 1 proposition – implying an expected development cost of $1.795 trillion dollars per drug.
How absurd are these figures on R&D costs? Nearly all current estimates of the costs of the development of new drugs are based upon a 1991 industry funded study by several academic economists who work with the drug industry.(6) According to this study, which used data supplied by the Pharmaceutical Manufactures Association (PMA), the average out-of-pocket costs of clinical trials for new drug approvals was $20.4 million. The industry made this number seem more impressive by adjusting the out-of-pocket costs for the “dry-hole” risks of failure and for the cost of capital. This gave the researchers a number of about $75 million. It took wild guesses about investments in the pre-clinical stage of research to get the number to $231 million, and none of this took into account who actually pays for the research. The U.S. government funds much of the pre-clinical research on drugs.
This 1991 study has subsequently become the basis for most estimates of drug development costs. The estimate took on a new life in 1993, when the U.S. Office of Technology Assessment (OTA) published a report on the costs of drug development.(7) Since OTA wasn’t able to obtain its own data on drug development costs, it hired Joseph DiMasi from Tufts University, one of the authors of the 1991 study sponsored by PMA, and recalculated his estimate using new assumptions for the cost of capital for drug development. To establish the “upper bound” for the cost of drug development, OTA assumed a 14 percent real rate of return(8) for the investments in the early pre-clinical research, and obtained a number of $359 million for the cost of developing a new drug. Now the industry had succeeded in getting its own analysis inserted into an official U.S. government report. Few persons understand how this number was developed, and what it represents.
Most important, few persons understand the difference between “out of pocket” costs, which are a tiny fraction of the final number, and the numbers which are adjusted for risk and capital costs. And, few persons understand that most of the final number is mostly based upon heroic estimates of the costs of pre-clinical research, much of which is paid for by the government, and conducted in government and university laboratories.(9)]
There is one source of publicly available data on industry drug development costs from the United States which is particularly instructive. In the United States, there is a 50 percent tax credit for expenditures on clinical trials for Orphan drugs. Between 1989 and 1993, the drug companies claimed $86.6 million in Orphan Drug tax credits, for an implicit industry cost of clinical trials on Orphan Drugs of $173.2 million.(10) During that same 5 year period, the FDA approved for marketing 60 Orphan Drugs. This works out to $2.9 million per drug approved for marketing. Of course, not all drugs that are tested receive FDA marketing approval, so the $2.9 million number is adjusted for the “dry-hole” risk. These are interesting figures, and give lie to the more dramatic assertions by the industry regarding the costs of drug development.
In Argentina and in Brazil, I took the position that it would be futile to suggest that there would be no international agreements on pharmaceutical drug development. It is necessary to develop an intellectually sound framework for such agreements, which place public health considerations front and center. It is a fact that the U.S. and other developed countries will seek broader sharing of the costs of R&D for new drugs.
In our view, the first task is to frame the R&D question as a public health issue. Increased levels of R&D on pharmaceuticals are a good thing for public health. Policies which enhance the public’s access to new health care inventions are also a good thing for public health.’
Increased profits to pharmaceutical companies are not an end, but
rather one of several mechanisms which may or may not contribute to the
size and efficacy of the R&D effort. This is an important point.
We have proposed three basic ways that national policies can increase pharmaceutical R&D.
1. Award patents on new pharmaceutical inventions, thereby creating financial incentives for firms to develop new drugs. This is the focus of the current TRIPS accord and much of the U.S. government lobbying on this issue. There are obvious shortcomings and limits to a policy of awarding monopolies on important health care discoveries. Monopolies seek to charge high prices, and high prices prevent many consumers from obtaining access to the new technologies. This is a contradiction which will not disappear, and which will become even more obvious to citizens in developed countries as the new health care technologies are increasingly aggressively priced. When a new start-up company decides to charge more than $500,000 for a single year of Ceredase, a drug used to treat Gaucher’s disease, we can see the kinds of barriers to access that lie ahead. There is also the problem of research priorities. For example, in South America, there is very little research on Chagas’ disease, because most persons who suffer from this terrible illness are poor.
2. A country can fund research on health care directly from its own budget. Government funding of health care R&D has worked well in the United States. The government has focused research efforts on important public health issues. If the government wants to, it can place the inventions in the public domain, broadening access to the technology.(11)
3. The government can also require pharmaceutical companies to reinvest in health care R&D. The money can be invested directly by the companies themselves, it can be diverted to a national R&D fund, to be administered by public health authorities, or there can be a combination of the two. There are many advantages to the mandatory reinvestment approach. The country can determine the aggregate level of R&D, and this can be done without regard to the private sector’s views on patent policies, price controls or other areas where we frequently see a form of blackmail by the industry. It is, after all, the consumers who fund the pharmaceutical companies, and this approach simply guarantees that an acceptable amount of the revenues from drug sales are actually spent on R&D. This is not simply an idea for developing countries, although I have proposed this in Argentina and Brazil. It is an idea that we are also trying to promote in the United States.
In our experience, efforts to protect consumers from high prices for pharmaceutical drugs are resisted on the grounds that price controls, compulsory licensing or other practices lead to reduced R&D investments. Patient groups are often cynically manipulated by these threats. This is a way out of that dilemma. This is a way to reconcile the public health interest in the development of new therapies with the broadest access to the new technologies.
The initial proposal for a mandatory reinvestment requirement was made by a pharmaceutical company, Andrulis, when it was seeking a non-exclusive license to sell cisplatin, a government funded cancer drug, which was marketed by Bristol-Myers. Bristol-Myers said if it could not make large profits from the exclusive license it would not be able to invest in R&D on new drugs. The Bristol-Myers argument was specious, since the profits on a drug already developed by the government is not a forward looking incentive. But the Andrulis counter proposal was an important innovation.
Andrulis asked the government to provide non-exclusive licensing of the drug, but to require each license holder to make contributions to a fund for R&D. Andrulis pointed out that the government could set any level of contribution it wanted, making the level of R&D investment a variable completely under the control of the government.
On September 27, 1996, Representative Bernie Sanders (R-VT) introduced legislation in the U.S. Congress (HR 4270, 104th Congress) that would require a minimum level of R&D reinvestment for each drug sold in the United States.(12) The Sanders proposal, which we support, would require each firm to create a fund for R&D, based on a percentage of the company’s sales. The contribution levels would depend on patent protection, orphan drug status, and the magnitude of sales.(13) The investments themselves would be made by the companies, so the government would make the macro decision regarding the minimum level of R&D investment, and the companies would make the micro decisions regarding the specific projects to receive funding. The Sanders bill did not pass this year, but it will be reintroduced in the next Congress.
Our proposal for a new framework for trade agreements on pharmaceuticals focuses on international efforts to share the costs of R&D, while leaving each country the flexibility to choose the particular mechanisms to promote R&D. Any combinations of the options described above would be acceptable. The level of expected effort should depend upon measures of ability to pay, such as per-capita income.
Finally, I would like to encourage this audience to consider the importance of better collection of economic data on pharmaceutical development costs. The industry has far too tight a control over statistics on drug development costs, and they use this control to deter independent research on pharmaceutical policies. The Sanders proposal (HR 4270, 104th Congress) outlines many of the statistics that should be collected, and we hope that organizations such as Health Care International will support such data collection efforts.
Thank you for the opportunity to appear today.
2 Until 1991, the FDA rates drugs on the basis of efficacy and the severity of the illness. Our sample included all FDA “A,” “AA” and “E” drugs. These were the codes used by the FDA to identify drugs that offered significant improvements over existing therapies, drugs to treat AIDS, or drugs for the treatment of severe illnesses. In 1992 the FDA eliminated it’s efficacy rating.
3 James Love, “The Orphan Drug Act and Government Sponsored Monopolies for Marketing Pharmaceutical Drugs,” Comments Submitted to the Subcommittee on Antitrust, Monopolies and Business Rights of the Committee on the Judiciary, U.S. Senate. January 21, 1992.
5 James Love, “Pharmaceutical Drugs, Intellectual Property Rights and Public Health: A Consumer Perspective from the United States,” Presented at XV Asamblea General de la Asociaión Latinoamericana de Industrias Farmacéuticas, San Carlos de Bariloche – Río Negro – Argentina, 11 al 13 de mayo de 1994., Remarks delivered on May 12, 1994..
6 DiMasi, Hansen, Grabowski and Lasagna, “Cost of innovation in the pharmaceutical industry, “Journal of Health Economics, 10, 1991, pages 107-142.
7 Pharmaceutical R&D: Costs, Risks and Rewards, OTA-H-522, February 1993.
8 14 percent plus the rate of inflation.
9 The researchers did not have project level data for pre-clinical research.
10 The most recent 5 year period for which data exist.
11 The United States government often assigns exclusive rights to government funded inventions, even when the commercial marketing of the drug would take place without the assigned of exclusive rights.
13 Section 7.