F or several decades, the pharmaceutical industry has benefited from a combination of government intervention and laissez faire: the federal government provides stringent intellectual property right protections and generous public subsidies for research-- but does not regulate drug prices. As a result, the United States has been a leader in the development of new drugs, but it also faces the highest drug prices in the world.
Until recently, there has been little public controversy over the pricing of drugs or the terms under which private firms obtain the rights to government-funded research. But as health care costs have soared and policy makers attempt to deal with AIDS and the general crisis of health coverage and cost, the drug companies are coming under increasing scrutiny.
No sooner had Bill Clinton taken office than he began criticizing drug prices. As part of national health care reform, the administration and Congress are considering various proposals to subsidize vaccines and broaden government and private insurance coverage of pharmaceutical products. Measures to expand health care coverage while containing costs would seem to require new attempts to control drug prices.
As part of national health care reform, the administration is rethinking how public interests intersect--and collide. An important part of this question is how the government manages the transfer of publicly funded drug research and to what extent it regulates private companies that benefit from this transfer. The stakes are enormous. If the drug companies are to be believed, government attempts to control drug prices would cripple the industry's research and development efforts, slow the pace of innovation, and damage one of the nation's leading high-technology export industries. But if the government fails to act, the result will be continued inflation of drug prices, often at the expense of patients who can't afford to buy drugs they need, taxpayers who bear costs under medicaid and medicare, and consumers who generally face higher insurance premiums.
T he federal government funds about 42 percent of all U.S. health care research and development (R&D) expenditures, including a significant portion of R&D costs for new drugs.The government plays a particularly important role in the highest risk research projects, including basic research, where commercial payoffs are least certain. It also pays a significant share of the later stages of drug development. For example, in the area of federal expenditures on human use clinical trials, a relatively advanced area for drug development, the National Institutes of Health (NIH) will spend an estimated $868.8 million in fiscal year 1993.
The Center for the Study of Responsive Law's Taxpayer Assets Project (TAP) recently studied the role of federal funding in the development of new drugs approved for marketing by the Food and Drug Administration between 1987 and 1991. The results illustrate the degree to which the drug industry relies on government money to develop privately owned products.
While the FDA approves hundreds of drugs every year, the number of new or important drugs is relatively small. For example, in 1991 the FDA approved 327 new and generic drugs and biologic products. Only 30 approvals were for new molecular entities (NMEs)--drugs distinctly different in structure from those already on the market. Only ten of these drugs received a priority rating, which is reserved for drugs that afford a "significant therapeutic gain," treat "severely debilitating or life threatening illness," or treat AIDS. For the group, seven of the ten priority drugs were developed with significant federal funds.
The Taxpayer Assets Project also studied the funding of all cancer drugs that were discovered since the National Cancer Institute's (NCI) new drug program began in 1955 and approved for marketing by the FDA through 1992. Of the 37 cancer drugs, 92 percent, or 34 cancer drugs, were developed with federal funding.
A more surprising finding of the Taxpayer Assets study concerned the pricing of the NMEs that received FDA approval from 1987 to 1991. The median wholesale cost (a completed treatment or a year, whichever was less) was $1,485 for the drugs that were developed without federal funding, and $4,480 for the drugs that were developed with federal funding. That is, the drugs that were developed with government funding were 3 times as expensive as the drugs developed without government funding. In 1991, the most recent year of the study, drugs developed with federal funding were over 11 times more expensive than drugs developed without federal funding.
Why are drugs developed with government funding so expensive? From the point of view of the drug companies, the answer is why not--that is, why not charge whatever the market will bear?
THE COST OF DRUG DEVELOPMENT
Drug companies emphasize the high costs and risks associated with the development of new drugs and argue that these factors justify the policies concerning the transfer of government-funded technology to the private sector. However, while there is broad recognition that drug development is a risky and costly enterprise, there is considerable controversy over the methods used to "transfer" ownership of government funded technology to the industry.
The most widely quoted estimate for the cost of developing a new drug was a Tufts University study sponsored by the pharmaceutical industry, which pegged the average cost of developing a new drug at $231 million, based on industry data on the costs of clinical trials for 99 new drugs.
The Tufts study's $231 million dollar figure has been widely misinterpreted. The Tufts researchers found that the average inflation-adjusted cost of clinical research was about $20.4 million. But by including a number of adjustments, including the "dry hole" risks of failures and the opportunity costs of capital (foregone profits) the researchers came up with a figure of approximately $75 million. To get the $231 million figure, the Tufts researchers added $156 million, which they estimated to be the cost of preclinical research, adjusted for inflation, the cost of capital, and the risk of failure. The $156 million for preclinical research, however, was not supported by project level data, but was calculated on the basis of very rough aggregated data, using heroic assumptions.
While the industry has used the Tufts study to emphasize the high costs of drug development, it can also be used to argue that the prices for drugs developed with federal funds should be priced much lower than drugs developed without federal funding. If, for example, the government has funded the preclinical research, then two-thirds of the cost of developing a new drug has already been paid for. And if the drug company obtains the rights to the drug after the conclusion of Phase II trials, more than 84 percent of the development costs are already accounted for.
The Tufts study also dramatically illustrates the significance of the point at which a company acquires the technology. Government-funded medical R&D typically focuses on the early stages of a drug's development, when the risks are the highest. For many drugs, the government has paid for most or all of the preclinical research, and it frequently funds the development of the drug all the way through FDA Phase II and Phase III trials. In these cases, which are many, the drug should not be priced as though the firm had borne all the risks and made all the investments. After all, citizens should not have to pay twice for the development of the drug, first as taxpayers and then as health care consumers. But current federal policies for managing what drug companies do with government research still reflect the priorities of the Reagan and Bush era agenda, where a commitment to corporate interests often came at the public's expense.
T hroughout the 1980s, lawmakers enacted a series of "technology transfer" laws designed to provide incentives for commercial development and to prevent foreign interests from benefiting from U.S.-funded research and development. These laws made it increasingly easy for drug companies to obtain exclusive rights to federal research without being subject to pricing controls. To appreciate how this combination of protection and laissez faire plays out, consider the case of the drug Taxol.
Taxol, an important new oncology drug, may be an effective treatment for breast, lung, and ovarian cancers. Taxol's only approved source is the bark of the Pacific Yew, a rare and slowly maturing tree that is found mostly on federal lands.
Taxol was discovered, manufactured, and tested in humans by NCI over a 30-year period. Early studies on cancer patients were carried out under government grants at a number of universities. By 1991 the federal government had completed Phase II clinical trials on six types of cancer, and had plans to test Taxol on 24 more.
According to Dr. Samuel Broder, Director of NCI, the federal agency was "totally responsible" for the development of Taxol, including the collection of the Yew bark; all biological screening in both cell cultures and animal tumor systems; chemical purification, isolation, and identification; and sponsorship of all clinical trials. Broder has estimated the taxpayers will spend about $35 million on past and future Taxol research.
Rather than allow many firms to develop Taxol competitively, NCI decided to award the rights to a single firm in the form of a CRADA (a Cooperative Research and Development Agreement, a contract between federal agencies and firms outlining the terms of joint research efforts). The notice for the CRADA was published in the Federal Register in August 1989, and firms were given just 45 days to respond, despite the complexity of the CRADA proposal. Four companies responded. The winning "bidder," was Bristol-Myers, a firm that was particularly well prepared, due largely to the fact that it had hired an NCI official, Dr. Robert Wittes, who had knowledge of the NCI Taxol program. The Bristol-Myers Squibb "bid" was submitted jointly with Hauser Chemical company, the firm that was then under contract to NCI to manufacture Taxol for the government's clinical trials.
The Bristol-Myers/NCI CRADA gave the firm exclusive rights to NCI's government-funded research, including the records of research completed before Bristol-Myers entered the Taxol picture, as well as all "new studies and raw data" from future NCI-funded Taxol research, which NCI agreed to make "available exclusively to Bristol-Myers," so long as the company is "engaged in the commercial development and marketing of Taxol." The company also received the exclusive rights to harvest the Pacific Yew trees found on federal lands.
In return, the government receives no money or royalties, but only Bristol-Myers Squibb's "best efforts" to commercialize Taxol, including a commitment to supply Taxol for government-run clinical trials, which were needed to obtain FDA marketing approval for the drug, and to an ambiguous "fair pricing" clause for Taxol
The fair pricing clause is dubious. Prior to the CRADA, NCI had used Hauser Chemical, a private firm, to manufacture Taxol for research purposes. Bristol-Myers Squibb continued to contract with Hauser both to supply the government with approximately 17 kilos of Taxol and to provide Bristol-Myers Squibb Taxol for commercial sales, once the company received FDA marketing approval. According to Securities and Exchange Commission filings, Hauser agreed to supply Bristol-Myers Squibb with more than 400 kilos of Taxol by August of 1994, subject to FDA marketing approval, for approximately $100 million, or about $.25 per milligram.
The FDA approved Taxol for sale in the United States in December of 1992, and Bristol-Myers Squibb announced a wholesale price of $4.87 per milligram, more than 19 times the cost of the drug from Hauser. To appreciate the magnitude of the markup, consider that the 400 kilos of Taxol produced by Hauser for $100 million had a wholesale value of $1.94 billion. (The difference between the cost of the drug from Hauser and the wholesale value of the product was greater than the entire cost of all drug company investments in human use clinical trials in 1989, the latest year for which data are available.)
For a patient taking Taxol, who responds to the treatment, the cost of the drug may exceed $10,000--while Bristol-Myers Squibb's costs of manufacturing the drug are about $500. Based upon the available data, it is unlikely that BMS spent more than $5 million manufacturing Taxol for NCI sponsored clinical trials prior to receiving FDA approval.
The company has defended its pricing of Taxol by making sweeping assertions of the huge investments that it has made to secure "future supplies" of Taxol or Taxol analogues. But these "investments" appear to consist primarily of "commitments" for long-term contracts, such as the Hauser contract, which are related to obtaining supplies for its commercial sales of Taxol, or to secure alternative sources of Taxol.
Thus while the invention of Taxol was in the public domain, and an important source of the drug was found on public lands, NCI was able to create substantial barriers that would discourage other firms from entering the Taxol market.The Taxol CRADA also illustrates the lengths to which the government will go to enhance monopoly power in the marketing of new drugs, even when technologies are not patentable. Taxol sales are expected to exceed $800 million per year, which are large, even by industry standards.
T he technology transfer acts of the 1980s have also made it easier for nonprofit institutions doing government-funded research to obtain property rights. Non-government organizations, however, have few incentives to manage R&D property rights in the public interest. In 1990, 84 percent of the NIH's $7.14 billion R&D budget wasused by nonprofit institutions, including $4.18 billion by universities and $1.8 billion by other nonprofit institutions. With billions of dollars at stake, universities and their faculties have pursued the licensing and marketing of new medical technologies, looking only at the potential marketing profits.
According to the National Science Board, academic patents on health and biomedical related inventions have increased particularly rapidly, constituting about 24 percent of all academic patents received in the late 1980s--double their share a decade ago. There is also considerable speculation that many important federally funded health care inventions are patented privately, by firms with ties to faculty members who want to avoid sharing royalties and licensing fees with their university employers.
There are also substantial problems concerning conflicts of interest. For example, the Scripps Research Institution has had a first right of refusal contract with Johnson & Johnson to commercialize certain chemical and pharmaceutical research, while several members of the Scripps faculty, including the president, have independent consulting agreements with Johnson & Johnson. The president of Scripps, who must negotiate the royalty and profit-sharing agreements between Johnson & Johnson and Scripps, is on the payroll at both institutions. Scripps has just signed a ten-year contract with Sandoz, the Swiss pharmaceutical firm, which gives this foreign-owned company the rights to commercialize an estimated billion dollars in federal health care research--a kind of reverse industrial policy.
PHARMACEUTICAL ORPHANAGES
In those cases where academic researchers do publish findings in journals and enter the public domain, pharmaceutical firms are able to obtain seven years of exclusive marketing rights under the provisions of the federal Orphan Drug Act, regardless of whether or not the company contributed to the research that led to the drug's discovery or knowledge of its efficacy in treating particular diseases. The law, originally enacted to encourage the development of drugs that would be unprofitable due to small markets, has been repeatedly modified to the advantage of pharmaceutical companies.
Today the only condition imposed on companies seeking orphan drug designation is that a drug must serve a client population of under 200,000. But this number is deceptive. For example, an estimated 6.8 million Americans suffer from cancer, but, a firm can distinguish particular types of cancer for orphan designation. Thus, for example, ovarian cancer, the fifth-leading cause of death among women victims of cancer, has an estimated client population of 164,000, well under the 200,000 limit.
Under the Orphan Drug Act, the FDA has become an ad hoc rival to the Patent and Trademark Office. While that office allocates exclusivity marketing rights to inventors, the FDA awards exclusivity to the first firm that obtains FDA marketing approval. In some instances, one orphan blocks entrance by other drugs that have their own patents and arguably different medical characteristics. For many drugs, the Orphan Drug Act actually adds a new element of risk to the development process, as it is possible to be barred from marketing a drug with a valid patent. Firms with patents have even been beaten to the punch by firms that don't hold patents, creating cases where the firm that holds the FDA orphan drug exclusivity must license the patent from the firm barred from the market.
The Orphan Drug Act has vastly increased the monopoly pricing power for many drugs, and it has created special challenges for drugs developed with public funds. The first firm to obtain FDA approval to market a drug that can qualify as an orphan is automatically granted marketing exclusivity, regardless of the company's role in the drug's development.
LITTLE ORPHAN CEREDASE
To appreciate how the orphan drug act has worked, consider the case of Ceredase (trade name for Algucerase), which is used to treat Gaucher disease, a severely debilitating disease that causes hematologic disorders, enlargement of the liver and spleen, bone erosion, and pain.
Ceredase and its efficacy in treating Gaucher disease were discovered by NIH researchers. Beginning in 1976, NIH entered into a series of contracts with the New England Enzyme Center at Tufts University to manufacture Ceredase for use in several clinic trials administered by NIH. In 1981 the New England Enzyme Center was closed, and the Ceredase contracts were transferred to Genzyme, a profit-making firm whose founders included Henry Blair, the former head of the Tufts Center. Through 1992, Genzyme received nearly $9 million from NIH, plus more than $5 million in fees from patients receiving the drug under an FDA investigational drug program.
In 1991 Genzyme received FDA approval to market Ceredase commercially, including seven years of exclusive marketing under the Orphan Drug Act. Despite NIH's role in the discovery and development of Ceredase and the grant of an orphan drug marketing monopoly, the government has no control over the prices the company charges.
The cost of the drug to patients, most of whom must take it indefinitely, is between $77,000 and $552,000 per year. (The first year of treatment is very costly, followed by lower maintenance costs.) In early 1992 Genzyme said it was treating more than 900 patients, who were paying an average of $140,000 per year for the drug. The high cost of treatment has imposed extreme hardships on patients, sometimes exhausting the lifetime benefits of private health care policies, leaving their families without insurance for other medical expenses.
The federal government does not require companies to disclose their costs, but based on figures provided to the Congressional Office of Technology Assessment, Genzyme claims that its costs of manufacturing, distributing, and marketing Ceredase were $1.60 per unit in 1992, or less than half of the $3.50 per unit cost of the drug. This includes charges to the depreciation of manufacturing facilities, which will apparently be used for products other than Ceredase. Genzyme claims bad debts and free distribution of the drug to indigent patients to cost about $.30 per unit, leaving a profit of $1.60 per unit. Manufacturing costs are expected to decline substantially as the drug is produced in larger quantities. Ceredase could generate a billion dollars in revenues for Genzyme over the seven years of its marketing monopoly.
Attempts to redirect the Orphan Drug Act toward its originally conceived mission have generally failed. In 1990, George Bush vetoed legislation that would have allowed the FDA to consider rates of growth of client groups, largely to remove AIDS from the orphan drug list, and which would have granted marketing rights to companies that develop drugs simultaneously. In 1992, legislation that would have eliminated orphan drug marketing exclusivity after the drug generated $200 million in cumulative revenues died in the face of stiff opposition from the drug companies, despite testimony that revenues for many "orphans" vastly exceeded development costs.
IS FAIR FAIR?
While the need for drug pricing controls has become clear to researchers and policymakers, federal attempts to implement controls have been halfhearted at best. The NIH instituted the "fair pricing" guidelines only in response to the public outrage over the $10,000 price tag Burroughs Wellcome put on the AIDS drug AZT. The drug had been largely developed by government money and researchers.
Officials at NIH are extremely uncomfortable with questions about the pricing of medical technology. The agency's primary mission is to promote the advancement of science in the health care field, and it has demonstrated neither the interest nor the expertise to develop useful models for setting prices. But even with the best of intentions, NIH is only one of several federal agencies managing health care R&D, and in every case agencies must work within a legal framework that increasingly promotes exclusive ownership of federally funded R&D.
NIH has used the fair pricing clause sparingly. The first two fair price agreements were for Taxol and a separate agreement with Bristol-Myers Squibb to market the AIDS drug ddI, a drug patented by NIH and licensed to Bristol-Myers Squibb on an exclusive basis for ten years.
Nothing in the Taxol fair pricing clause addresses the relationship between the company's investments, risks and, the product price--a deliberate omission. For both ddI and Taxol, NIH officials compared the company's announced price to the prices of products used to treat similar diseases. Thus, for example, the Bristol-Myers Squibb price for ddI was considered reasonable because it was less than the initial price charged for the drug AZT. The irony of using a comparison to the price for AZT, which was widely considered to be excessive, is obvious. NIH never asked Bristol-Myers Squibb to compare the price of the drug with its costs of manufacturing, testing, and marketing.
To appreciate the flaws in this approach, consider the initial NIH analysis of the "fair price" for Taxol. While NIH has yet to ask Bristol-Myers Squibb for data on the company's costs in manufacturing, testing, and marketing the drug, it is a matter of public record that Bristol-Myers Squibb obtains Taxol from its supplier for less than $.25 per milligram.
Rather than focus on Bristol-Myers Squibb's actual costs, NIH officials told the firmto price Taxol in the range of other cancer drugs. NIH submitted to Bristol-Myers Squibb a list of 15 drugs and their estimated "monthly wholesale cost." Bristol-Myers Squibb was asked to price Taxol so that a month of Taxol would cost no more than the median for the group.
NIH was, in essence, telling Bristol-Myers Squibb that it could price Taxol, a government-funded drug invention, the same as other cancer drugs were priced, regardless of where the funding came from. The median cost of the drugs on the list was $1,776. The BMS wholesale cost of Taxol for a typical patient suffering from ovarian cancer is about $1,685 per month--compared with a cost of $86.50 to produce the drug.
F ederal efforts to control the prices of drugs developed with federal funds is part of a larger problem of controlling drug prices. Senator David Pryor and Congressman Ron Wyden have been vocal critics of NIH's feeble efforts to introduce fair pricing agreements on federal government CRADA and license agreements, and NIH is reviewing its policies. But even if NIH adopts better methods of determining fair prices, most of the federal funding for new drug development is performed by universities and other institutions and therefore not under the control of any federal agency.
The most difficult problem is to set prices on new drugs just entering the market. The pharmaceutical industry proposes to limit the rate of overall drug price increases to the consumer price index in an attempt to avoid such price controls. The industry proposal will control only on the rate at which prices can be increased, not the roll out prices on new drugs. And older drugs are expected to face downward pressure on prices from managed competition, even without the industry's proposed voluntary limits on the rate of price increases.
T he Clinton administration has not yet made official statements regarding its proposals to control drug prices. In May the New York Times reported that the administration would ask for the creation of a drug review board, which would investigate the pricing of drugs, but would not set prices. The board's powers would be limited to moral suasion.
But a drug review board should be more than just the government's conscience. It should collect detailed information on industry costs and revenues and set prices or limit patent protection for drugs priced too high. It should also be required to consider the extent to which a drug's development costs were paid for by taxpayers.
Moreover, universities and other recipients of federal grants and contracts should be required to disclose the contents of contracts for exclusive development and marketing rights on government-funded drugs, and to allow the federal government to review the agreements, following public comment, and determine if the agreements are in the public interest.
While the most important factor driving the politics of drug pricing are the high powered fundraising, lobbying, and public relations efforts of the drug companies, who believe they are in the biggest battle of their lives, there are also important dilemmas facing policymakers regarding the impact of regulatory schemes of the industry's R&D investments.
In many cases, federal efforts to limit drug prices will lead to less private sector R&D investments. The current arrangement, however, is a particularly inefficient way to attract R&D investments. The federal government should collect detailed information on the actual investments in drug R&D, and to set national targets regarding investments. If R&D falls below target rates, the federal government has a number of options, including direct subsidies to firms (through tax credits or grants). One promising idea surfaced from the industry several years ago. When Bristol-Myers Squibb was seeking a renewal of its exclusive license of the cancer drug Cisplatin, several companies asked the government to make the drug available on a non-exclusive basis. One company proposed that the federal government impose a royalty on the non-exclusive license and then use the money to fund R&D. NCI rejected the proposal.
The prospects for reform on these matters are surprisingly remote. Some policymakers say it would be politically impossible even to collect data from the drug companies on drug prices and revenues, despite the fact that private sector firms, such as Dun and Bradstreet, already collect and disseminate these data for the industry. Congressional staffers consider it virtually unthinkable that Congress could muster the votes for any legislation that gives the government the power to regulate the prices for new drugs.
The political and technical problems inherent in controlling the prices of pharmaceutical drugs are similar to the problems that confront many new sectors of our increasingly sophisticated economy. Faced with dynamic changes in technology and a Congress and civil service that pander to industry pressures, there is a disheartening tendency to abandon efforts to protect consumers or taxpayers. If the public service is to gain the respect that it once enjoyed, however, policymakers have to confront these difficult problems and find ways to make the government work to the benefit of ordinary citizens.
F or several decades, the pharmaceutical industry has benefited from a combination of government intervention and laissez faire: the federal government provides stringent intellectual property right protections and generous public subsidies for research-- but does not regulate drug prices. As a result, the United States has been a leader in the development of new drugs, but it also faces the highest drug prices in the world.
Until recently, there has been little public controversy over the pricing of drugs or the terms under which private firms obtain the rights to government-funded research. But as health care costs have soared and policy makers attempt to deal with AIDS and the general crisis of health coverage and cost, the drug companies are coming under increasing scrutiny.
No sooner had Bill Clinton taken office than he began criticizing drug prices. As part of national health care reform, the administration and Congress are considering various proposals to subsidize vaccines and broaden government and private insurance coverage of pharmaceutical products. Measures to expand health care coverage while containing costs would seem to require new attempts to control drug prices.
As part of national health care reform, the administration is rethinking how public interests intersect--and collide. An important part of this question is how the government manages the transfer of publicly funded drug research and to what extent it regulates private companies that benefit from this transfer. The stakes are enormous. If the drug companies are to be believed, government attempts to control drug prices would cripple the industry's research and development efforts, slow the pace of innovation, and damage one of the nation's leading high-technology export industries. But if the government fails to act, the result will be continued inflation of drug prices, often at the expense of patients who can't afford to buy drugs they need, taxpayers who bear costs under medicaid and medicare, and consumers who generally face higher insurance premiums.
T he federal government funds about 42 percent of all U.S. health care research and development (R&D) expenditures, including a significant portion of R&D costs for new drugs.The government plays a particularly important role in the highest risk research projects, including basic research, where commercial payoffs are least certain. It also pays a significant share of the later stages of drug development. For example, in the area of federal expenditures on human use clinical trials, a relatively advanced area for drug development, the National Institutes of Health (NIH) will spend an estimated $868.8 million in fiscal year 1993.
The Center for the Study of Responsive Law's Taxpayer Assets Project (TAP) recently studied the role of federal funding in the development of new drugs approved for marketing by the Food and Drug Administration between 1987 and 1991. The results illustrate the degree to which the drug industry relies on government money to develop privately owned products.
While the FDA approves hundreds of drugs every year, the number of new or important drugs is relatively small. For example, in 1991 the FDA approved 327 new and generic drugs and biologic products. Only 30 approvals were for new molecular entities (NMEs)--drugs distinctly different in structure from those already on the market. Only ten of these drugs received a priority rating, which is reserved for drugs that afford a "significant therapeutic gain," treat "severely debilitating or life threatening illness," or treat AIDS. For the group, seven of the ten priority drugs were developed with significant federal funds.
The Taxpayer Assets Project also studied the funding of all cancer drugs that were discovered since the National Cancer Institute's (NCI) new drug program began in 1955 and approved for marketing by the FDA through 1992. Of the 37 cancer drugs, 92 percent, or 34 cancer drugs, were developed with federal funding.
A more surprising finding of the Taxpayer Assets study concerned the pricing of the NMEs that received FDA approval from 1987 to 1991. The median wholesale cost (a completed treatment or a year, whichever was less) was $1,485 for the drugs that were developed without federal funding, and $4,480 for the drugs that were developed with federal funding. That is, the drugs that were developed with government funding were 3 times as expensive as the drugs developed without government funding. In 1991, the most recent year of the study, drugs developed with federal funding were over 11 times more expensive than drugs developed without federal funding.
Why are drugs developed with government funding so expensive? From the point of view of the drug companies, the answer is why not--that is, why not charge whatever the market will bear?
THE COST OF DRUG DEVELOPMENT
Drug companies emphasize the high costs and risks associated with the development of new drugs and argue that these factors justify the policies concerning the transfer of government-funded technology to the private sector. However, while there is broad recognition that drug development is a risky and costly enterprise, there is considerable controversy over the methods used to "transfer" ownership of government funded technology to the industry.
The most widely quoted estimate for the cost of developing a new drug was a Tufts University study sponsored by the pharmaceutical industry, which pegged the average cost of developing a new drug at $231 million, based on industry data on the costs of clinical trials for 99 new drugs.
The Tufts study's $231 million dollar figure has been widely misinterpreted. The Tufts researchers found that the average inflation-adjusted cost of clinical research was about $20.4 million. But by including a number of adjustments, including the "dry hole" risks of failures and the opportunity costs of capital (foregone profits) the researchers came up with a figure of approximately $75 million. To get the $231 million figure, the Tufts researchers added $156 million, which they estimated to be the cost of preclinical research, adjusted for inflation, the cost of capital, and the risk of failure. The $156 million for preclinical research, however, was not supported by project level data, but was calculated on the basis of very rough aggregated data, using heroic assumptions.
While the industry has used the Tufts study to emphasize the high costs of drug development, it can also be used to argue that the prices for drugs developed with federal funds should be priced much lower than drugs developed without federal funding. If, for example, the government has funded the preclinical research, then two-thirds of the cost of developing a new drug has already been paid for. And if the drug company obtains the rights to the drug after the conclusion of Phase II trials, more than 84 percent of the development costs are already accounted for.
The Tufts study also dramatically illustrates the significance of the point at which a company acquires the technology. Government-funded medical R&D typically focuses on the early stages of a drug's development, when the risks are the highest. For many drugs, the government has paid for most or all of the preclinical research, and it frequently funds the development of the drug all the way through FDA Phase II and Phase III trials. In these cases, which are many, the drug should not be priced as though the firm had borne all the risks and made all the investments. After all, citizens should not have to pay twice for the development of the drug, first as taxpayers and then as health care consumers. But current federal policies for managing what drug companies do with government research still reflect the priorities of the Reagan and Bush era agenda, where a commitment to corporate interests often came at the public's expense.
T hroughout the 1980s, lawmakers enacted a series of "technology transfer" laws designed to provide incentives for commercial development and to prevent foreign interests from benefiting from U.S.-funded research and development. These laws made it increasingly easy for drug companies to obtain exclusive rights to federal research without being subject to pricing controls. To appreciate how this combination of protection and laissez faire plays out, consider the case of the drug Taxol.
Taxol, an important new oncology drug, may be an effective treatment for breast, lung, and ovarian cancers. Taxol's only approved source is the bark of the Pacific Yew, a rare and slowly maturing tree that is found mostly on federal lands.
Taxol was discovered, manufactured, and tested in humans by NCI over a 30-year period. Early studies on cancer patients were carried out under government grants at a number of universities. By 1991 the federal government had completed Phase II clinical trials on six types of cancer, and had plans to test Taxol on 24 more.
According to Dr. Samuel Broder, Director of NCI, the federal agency was "totally responsible" for the development of Taxol, including the collection of the Yew bark; all biological screening in both cell cultures and animal tumor systems; chemical purification, isolation, and identification; and sponsorship of all clinical trials. Broder has estimated the taxpayers will spend about $35 million on past and future Taxol research.
Rather than allow many firms to develop Taxol competitively, NCI decided to award the rights to a single firm in the form of a CRADA (a Cooperative Research and Development Agreement, a contract between federal agencies and firms outlining the terms of joint research efforts). The notice for the CRADA was published in the Federal Register in August 1989, and firms were given just 45 days to respond, despite the complexity of the CRADA proposal. Four companies responded. The winning "bidder," was Bristol-Myers, a firm that was particularly well prepared, due largely to the fact that it had hired an NCI official, Dr. Robert Wittes, who had knowledge of the NCI Taxol program. The Bristol-Myers Squibb "bid" was submitted jointly with Hauser Chemical company, the firm that was then under contract to NCI to manufacture Taxol for the government's clinical trials.
The Bristol-Myers/NCI CRADA gave the firm exclusive rights to NCI's government-funded research, including the records of research completed before Bristol-Myers entered the Taxol picture, as well as all "new studies and raw data" from future NCI-funded Taxol research, which NCI agreed to make "available exclusively to Bristol-Myers," so long as the company is "engaged in the commercial development and marketing of Taxol." The company also received the exclusive rights to harvest the Pacific Yew trees found on federal lands.
In return, the government receives no money or royalties, but only Bristol-Myers Squibb's "best efforts" to commercialize Taxol, including a commitment to supply Taxol for government-run clinical trials, which were needed to obtain FDA marketing approval for the drug, and to an ambiguous "fair pricing" clause for Taxol
The fair pricing clause is dubious. Prior to the CRADA, NCI had used Hauser Chemical, a private firm, to manufacture Taxol for research purposes. Bristol-Myers Squibb continued to contract with Hauser both to supply the government with approximately 17 kilos of Taxol and to provide Bristol-Myers Squibb Taxol for commercial sales, once the company received FDA marketing approval. According to Securities and Exchange Commission filings, Hauser agreed to supply Bristol-Myers Squibb with more than 400 kilos of Taxol by August of 1994, subject to FDA marketing approval, for approximately $100 million, or about $.25 per milligram.
The FDA approved Taxol for sale in the United States in December of 1992, and Bristol-Myers Squibb announced a wholesale price of $4.87 per milligram, more than 19 times the cost of the drug from Hauser. To appreciate the magnitude of the markup, consider that the 400 kilos of Taxol produced by Hauser for $100 million had a wholesale value of $1.94 billion. (The difference between the cost of the drug from Hauser and the wholesale value of the product was greater than the entire cost of all drug company investments in human use clinical trials in 1989, the latest year for which data are available.)
For a patient taking Taxol, who responds to the treatment, the cost of the drug may exceed $10,000--while Bristol-Myers Squibb's costs of manufacturing the drug are about $500. Based upon the available data, it is unlikely that BMS spent more than $5 million manufacturing Taxol for NCI sponsored clinical trials prior to receiving FDA approval.
The company has defended its pricing of Taxol by making sweeping assertions of the huge investments that it has made to secure "future supplies" of Taxol or Taxol analogues. But these "investments" appear to consist primarily of "commitments" for long-term contracts, such as the Hauser contract, which are related to obtaining supplies for its commercial sales of Taxol, or to secure alternative sources of Taxol.
Thus while the invention of Taxol was in the public domain, and an important source of the drug was found on public lands, NCI was able to create substantial barriers that would discourage other firms from entering the Taxol market.The Taxol CRADA also illustrates the lengths to which the government will go to enhance monopoly power in the marketing of new drugs, even when technologies are not patentable. Taxol sales are expected to exceed $800 million per year, which are large, even by industry standards.
T he technology transfer acts of the 1980s have also made it easier for nonprofit institutions doing government-funded research to obtain property rights. Non-government organizations, however, have few incentives to manage R&D property rights in the public interest. In 1990, 84 percent of the NIH's $7.14 billion R&D budget wasused by nonprofit institutions, including $4.18 billion by universities and $1.8 billion by other nonprofit institutions. With billions of dollars at stake, universities and their faculties have pursued the licensing and marketing of new medical technologies, looking only at the potential marketing profits.
According to the National Science Board, academic patents on health and biomedical related inventions have increased particularly rapidly, constituting about 24 percent of all academic patents received in the late 1980s--double their share a decade ago. There is also considerable speculation that many important federally funded health care inventions are patented privately, by firms with ties to faculty members who want to avoid sharing royalties and licensing fees with their university employers.
There are also substantial problems concerning conflicts of interest. For example, the Scripps Research Institution has had a first right of refusal contract with Johnson & Johnson to commercialize certain chemical and pharmaceutical research, while several members of the Scripps faculty, including the president, have independent consulting agreements with Johnson & Johnson. The president of Scripps, who must negotiate the royalty and profit-sharing agreements between Johnson & Johnson and Scripps, is on the payroll at both institutions. Scripps has just signed a ten-year contract with Sandoz, the Swiss pharmaceutical firm, which gives this foreign-owned company the rights to commercialize an estimated billion dollars in federal health care research--a kind of reverse industrial policy.
PHARMACEUTICAL ORPHANAGES
In those cases where academic researchers do publish findings in journals and enter the public domain, pharmaceutical firms are able to obtain seven years of exclusive marketing rights under the provisions of the federal Orphan Drug Act, regardless of whether or not the company contributed to the research that led to the drug's discovery or knowledge of its efficacy in treating particular diseases. The law, originally enacted to encourage the development of drugs that would be unprofitable due to small markets, has been repeatedly modified to the advantage of pharmaceutical companies.
Today the only condition imposed on companies seeking orphan drug designation is that a drug must serve a client population of under 200,000. But this number is deceptive. For example, an estimated 6.8 million Americans suffer from cancer, but, a firm can distinguish particular types of cancer for orphan designation. Thus, for example, ovarian cancer, the fifth-leading cause of death among women victims of cancer, has an estimated client population of 164,000, well under the 200,000 limit.
Under the Orphan Drug Act, the FDA has become an ad hoc rival to the Patent and Trademark Office. While that office allocates exclusivity marketing rights to inventors, the FDA awards exclusivity to the first firm that obtains FDA marketing approval. In some instances, one orphan blocks entrance by other drugs that have their own patents and arguably different medical characteristics. For many drugs, the Orphan Drug Act actually adds a new element of risk to the development process, as it is possible to be barred from marketing a drug with a valid patent. Firms with patents have even been beaten to the punch by firms that don't hold patents, creating cases where the firm that holds the FDA orphan drug exclusivity must license the patent from the firm barred from the market.
The Orphan Drug Act has vastly increased the monopoly pricing power for many drugs, and it has created special challenges for drugs developed with public funds. The first firm to obtain FDA approval to market a drug that can qualify as an orphan is automatically granted marketing exclusivity, regardless of the company's role in the drug's development.
LITTLE ORPHAN CEREDASE
To appreciate how the orphan drug act has worked, consider the case of Ceredase (trade name for Algucerase), which is used to treat Gaucher disease, a severely debilitating disease that causes hematologic disorders, enlargement of the liver and spleen, bone erosion, and pain.
Ceredase and its efficacy in treating Gaucher disease were discovered by NIH researchers. Beginning in 1976, NIH entered into a series of contracts with the New England Enzyme Center at Tufts University to manufacture Ceredase for use in several clinic trials administered by NIH. In 1981 the New England Enzyme Center was closed, and the Ceredase contracts were transferred to Genzyme, a profit-making firm whose founders included Henry Blair, the former head of the Tufts Center. Through 1992, Genzyme received nearly $9 million from NIH, plus more than $5 million in fees from patients receiving the drug under an FDA investigational drug program.
In 1991 Genzyme received FDA approval to market Ceredase commercially, including seven years of exclusive marketing under the Orphan Drug Act. Despite NIH's role in the discovery and development of Ceredase and the grant of an orphan drug marketing monopoly, the government has no control over the prices the company charges.
The cost of the drug to patients, most of whom must take it indefinitely, is between $77,000 and $552,000 per year. (The first year of treatment is very costly, followed by lower maintenance costs.) In early 1992 Genzyme said it was treating more than 900 patients, who were paying an average of $140,000 per year for the drug. The high cost of treatment has imposed extreme hardships on patients, sometimes exhausting the lifetime benefits of private health care policies, leaving their families without insurance for other medical expenses.
The federal government does not require companies to disclose their costs, but based on figures provided to the Congressional Office of Technology Assessment, Genzyme claims that its costs of manufacturing, distributing, and marketing Ceredase were $1.60 per unit in 1992, or less than half of the $3.50 per unit cost of the drug. This includes charges to the depreciation of manufacturing facilities, which will apparently be used for products other than Ceredase. Genzyme claims bad debts and free distribution of the drug to indigent patients to cost about $.30 per unit, leaving a profit of $1.60 per unit. Manufacturing costs are expected to decline substantially as the drug is produced in larger quantities. Ceredase could generate a billion dollars in revenues for Genzyme over the seven years of its marketing monopoly.
Attempts to redirect the Orphan Drug Act toward its originally conceived mission have generally failed. In 1990, George Bush vetoed legislation that would have allowed the FDA to consider rates of growth of client groups, largely to remove AIDS from the orphan drug list, and which would have granted marketing rights to companies that develop drugs simultaneously. In 1992, legislation that would have eliminated orphan drug marketing exclusivity after the drug generated $200 million in cumulative revenues died in the face of stiff opposition from the drug companies, despite testimony that revenues for many "orphans" vastly exceeded development costs.
IS FAIR FAIR?
While the need for drug pricing controls has become clear to researchers and policymakers, federal attempts to implement controls have been halfhearted at best. The NIH instituted the "fair pricing" guidelines only in response to the public outrage over the $10,000 price tag Burroughs Wellcome put on the AIDS drug AZT. The drug had been largely developed by government money and researchers.
Officials at NIH are extremely uncomfortable with questions about the pricing of medical technology. The agency's primary mission is to promote the advancement of science in the health care field, and it has demonstrated neither the interest nor the expertise to develop useful models for setting prices. But even with the best of intentions, NIH is only one of several federal agencies managing health care R&D, and in every case agencies must work within a legal framework that increasingly promotes exclusive ownership of federally funded R&D.
NIH has used the fair pricing clause sparingly. The first two fair price agreements were for Taxol and a separate agreement with Bristol-Myers Squibb to market the AIDS drug ddI, a drug patented by NIH and licensed to Bristol-Myers Squibb on an exclusive basis for ten years.
Nothing in the Taxol fair pricing clause addresses the relationship between the company's investments, risks and, the product price--a deliberate omission. For both ddI and Taxol, NIH officials compared the company's announced price to the prices of products used to treat similar diseases. Thus, for example, the Bristol-Myers Squibb price for ddI was considered reasonable because it was less than the initial price charged for the drug AZT. The irony of using a comparison to the price for AZT, which was widely considered to be excessive, is obvious. NIH never asked Bristol-Myers Squibb to compare the price of the drug with its costs of manufacturing, testing, and marketing.
To appreciate the flaws in this approach, consider the initial NIH analysis of the "fair price" for Taxol. While NIH has yet to ask Bristol-Myers Squibb for data on the company's costs in manufacturing, testing, and marketing the drug, it is a matter of public record that Bristol-Myers Squibb obtains Taxol from its supplier for less than $.25 per milligram.
Rather than focus on Bristol-Myers Squibb's actual costs, NIH officials told the firmto price Taxol in the range of other cancer drugs. NIH submitted to Bristol-Myers Squibb a list of 15 drugs and their estimated "monthly wholesale cost." Bristol-Myers Squibb was asked to price Taxol so that a month of Taxol would cost no more than the median for the group.
NIH was, in essence, telling Bristol-Myers Squibb that it could price Taxol, a government-funded drug invention, the same as other cancer drugs were priced, regardless of where the funding came from. The median cost of the drugs on the list was $1,776. The BMS wholesale cost of Taxol for a typical patient suffering from ovarian cancer is about $1,685 per month--compared with a cost of $86.50 to produce the drug.
F ederal efforts to control the prices of drugs developed with federal funds is part of a larger problem of controlling drug prices. Senator David Pryor and Congressman Ron Wyden have been vocal critics of NIH's feeble efforts to introduce fair pricing agreements on federal government CRADA and license agreements, and NIH is reviewing its policies. But even if NIH adopts better methods of determining fair prices, most of the federal funding for new drug development is performed by universities and other institutions and therefore not under the control of any federal agency.
The most difficult problem is to set prices on new drugs just entering the market. The pharmaceutical industry proposes to limit the rate of overall drug price increases to the consumer price index in an attempt to avoid such price controls. The industry proposal will control only on the rate at which prices can be increased, not the roll out prices on new drugs. And older drugs are expected to face downward pressure on prices from managed competition, even without the industry's proposed voluntary limits on the rate of price increases.
T he Clinton administration has not yet made official statements regarding its proposals to control drug prices. In May the New York Times reported that the administration would ask for the creation of a drug review board, which would investigate the pricing of drugs, but would not set prices. The board's powers would be limited to moral suasion.
But a drug review board should be more than just the government's conscience. It should collect detailed information on industry costs and revenues and set prices or limit patent protection for drugs priced too high. It should also be required to consider the extent to which a drug's development costs were paid for by taxpayers.
Moreover, universities and other recipients of federal grants and contracts should be required to disclose the contents of contracts for exclusive development and marketing rights on government-funded drugs, and to allow the federal government to review the agreements, following public comment, and determine if the agreements are in the public interest.
While the most important factor driving the politics of drug pricing are the high powered fundraising, lobbying, and public relations efforts of the drug companies, who believe they are in the biggest battle of their lives, there are also important dilemmas facing policymakers regarding the impact of regulatory schemes of the industry's R&D investments.
In many cases, federal efforts to limit drug prices will lead to less private sector R&D investments. The current arrangement, however, is a particularly inefficient way to attract R&D investments. The federal government should collect detailed information on the actual investments in drug R&D, and to set national targets regarding investments. If R&D falls below target rates, the federal government has a number of options, including direct subsidies to firms (through tax credits or grants). One promising idea surfaced from the industry several years ago. When Bristol-Myers Squibb was seeking a renewal of its exclusive license of the cancer drug Cisplatin, several companies asked the government to make the drug available on a non-exclusive basis. One company proposed that the federal government impose a royalty on the non-exclusive license and then use the money to fund R&D. NCI rejected the proposal.
The prospects for reform on these matters are surprisingly remote. Some policymakers say it would be politically impossible even to collect data from the drug companies on drug prices and revenues, despite the fact that private sector firms, such as Dun and Bradstreet, already collect and disseminate these data for the industry. Congressional staffers consider it virtually unthinkable that Congress could muster the votes for any legislation that gives the government the power to regulate the prices for new drugs.
The political and technical problems inherent in controlling the prices of pharmaceutical drugs are similar to the problems that confront many new sectors of our increasingly sophisticated economy. Faced with dynamic changes in technology and a Congress and civil service that pander to industry pressures, there is a disheartening tendency to abandon efforts to protect consumers or taxpayers. If the public service is to gain the respect that it once enjoyed, however, policymakers have to confront these difficult problems and find ways to make the government work to the benefit of ordinary citizens.
