Regulation, Litigation, and Innovation in the Pharmaceutical Industry: An Equation for Safety
MARVIN E. JAFFE, M.D
In 1992 Andrew Wiles of Princeton University demonstrated that there is no answer to the centuries-old theorem that French mathematician Pierre de Fermat left dangling in the margin of a notebook: x3 + y3 = z3. Indeed, it represents an "impossible triangle." Perhaps now he has time to direct his analytical powers toward another scientific dilemma: Does the threat of future liability restrict innovation in the pharmaceutical industry?
Unraveling the impact of the tort liability system on innovation in America is a difficult task for any industry, whether it is aviation, toys, or soap. But pharmaceuticals present an especially complex challenge. Like Fermat's equation, the question itself is unanswerable unless reference is made to the impact of the federal regulatory process that governs every aspect of the prescription drug cycle and makes reduction of liability risk intrinsic to new drug development. The effects of the tort liability system on innovation in pharmaceuticals cannot be calculated without factoring the pervasive role of the U.S. Food and Drug Administration (FDA) into the equation.
What is the relationship, therefore, between regulation, liability exposure, and pharmaceutical innovation? Are they three sides of an impossible triangle? To the contrary, the regulatory process helps create an environment where exposure to liability is far less a factor in pharmaceutical innovation than human and medical benefit are. Consider the fact that research is ongoing into the development of new products in challenging categories such as birth control, or that thalidomide, perhaps the most infamous
drug in history, is currently being "rehabilitated" for use against such devastating diseases as leprosy and AIDS.
CONSTRAINTS DESPITE SAFETY NET
Despite the regulatory safety net, however, the threat of litigation imposes constraints. Persistent, sometimes frivolous litigation casts a shadow on certain critical medical categories, notably vaccines. Even when no adverse judgments are made, companies bear the heavy costs of litigation and must pay high insurance premiums—or, most likely, self-insure. Overdesigned to reduce risk, the regulatory process allows little opportunity for serendipitous discovery once compounds enter the pipeline. Regulation also exacts a heavy toll on innovation through high costs and a plodding pace.
The asymmetrical system that arises from the divergent goals of drug regulation and the law of product liability is another key industry consideration. This paper explores this asymmetrical system and how it sets pharmaceuticals apart.
MEDICINE AND POISON
The most singular feature of the pharmaceutical industry is its products. Pharmaceuticals are designed for an intended medicinal effect, but as complex products, they may have unintended effects as well. In fact, the term "pharmaceutical" derives from the ancient Greek pharmakon , which literally means both "medicine" and "poison." Pharmaceuticals are categorized by the legal community as belonging to the class of "unavoidably unsafe products" including vaccines, blood, and medical devices, which offer desired benefits but are not without risk.1 The law recognizes that the medical value of pharmaceutical products differentiates them from other products such as lawnmowers or household cleaning products. The law characterizes drugs as ''not unreasonably dangerous because they are incapable of being made safe for their intended purpose."
INDUSTRY SPECIFIC SAFEGUARDS
While the law recognizes pharmaceuticals as "not unreasonably dangerous," the regulatory process is designed to protect consumers from excessive risk. No other industry in the United States has such extensive government oversight of the testing, formulation, manufacture, marketing, and distribution of their products. No other country has an equivalent regulatory system in terms of the extent of its control.
Drug regulation in the United States is pursued through several different avenues. New drugs must receive FDA approval before they can be marketed in the United States. The clinical research process is subject to direct FDA monitoring. Product labeling and promotional materials require regulatory approval as a prerequisite for sales.
Another distinguishing feature of the industry is the indirect way prescription drugs are sold. Pharmaceuticals are prescribed and dispensed by physicians and health care providers, or "learned intermediaries." Manufacturers bear the responsibility to educate the professional community about the risks and proper uses of products. In fact, manufacturers have the continuous legal obligation to "utilize methods of warning which will be reasonably effective"2 and are "required to keep abreast of the current state of knowledge of its products as gained through research, adverse reaction reports, scientific literature, and other available knowledge" (Fern and Sichel, 1985). This obligation to warn leads some industry experts to believe that product labeling is the key factor in reduction of pharmaceutical liability risk.
Reliance on learned intermediaries also reduces risk since the ultimate decision to select and prescribe a prescription drug lies with the physician and not the patient. While permitted, advertising directed to either patients or physicians is tightly regulated and scrutinized. A drug may not be promoted to the extent that an "otherwise adequate warning becomes inadequate."3 There are rigorous regulatory guidelines for packaging, promotional activities and events, educational forums, and even the sponsorship of research.
EMPOWERMENT OF THE FDA
The strict regulatory controls in place today evolved from two watershed events: deaths caused by lethal formulations of Elixir Sulfanilimide in the 1920s and tragic birth defects linked to the use of thalidomide in pregnant women in the 1960s. One hundred people died as a result of the use of diethylene glycol as a solvent for Elixir Sulfanilimide. Public reaction to this disaster led Congress to enact a "new drug" section in the 1938 Food, Drug, and Cosmetic Act. This empowered the FDA to evaluate the efficacy of all new drug formulations and to approve them as "safe for use" for indications specified on product labeling (Swazey, 1991).
Although never marketed in the United States, thalidomide was a popular sleeping drug introduced in 1957. It eventually was sold in 46 countries. In 1961 researchers discovered the association between thalidomide and phocomelia, or seal limbs, and other extreme congenital defects. By the end of 1961, thalidomide was withdrawn from most world markets. In the United States, the thalidomide tragedy led to the 1962 passage of the
Kefauver-Harris amendments to the Food, Drug, and Cosmetic Act, which tightened regulatory control over the safety of new drugs (Lasagna, 1991).
THE $359 MILLION ROAD TO MARKET
Today the drug development process typically spans 12 years and is costly and complex. Regulated development phases encompass laboratory testing; clinical studies of the pharmacologic profile of a new drug, its efficacy and tolerability by patients; and extensive clinical trials to study the effects of the drug in humans over specified periods of time. Pharmaceutical developers spend an average $359 million to bring a new drug to market, often for one limited application (U.S. Congress, Office of Technology Assessment, 1993). Very few drugs make the grade. In 1990 only 23 new drugs obtained clearance for marketing (Pharmaceutical Manufacturers Association, 1991). The attrition rate for new compounds is extraordinary. The Pharmaceutical Manufacturers Association estimates that only one in 5,000 new compounds completes the journey through the pipeline.
The drug development process begins with laboratory, or preclinical, testing of compounds that were discovered or acquired by a pharmaceutical manufacturer. During this time, which takes an average of three and one-half years, researchers seek to determine whether a compound is biologically active as well as safe. If preclinical tests yield promise in terms of human therapeutics, the pharmaceutical developer files an Investigational New Drug Application (IND) with the FDA before initiating clinical testing in volunteer patients. The IND provides exhaustive detail about the chemical, pharmacological, pharmaceutical, and toxicological properties of a new drug in the form in which it will be administered to patients.
Once the IND has been filed, approximately three years are spent in two phases of initial human clinical testing. In Phase I, 20 to 100 healthy people participate in studies where researchers observe the pharmacologic actions of the drug. These actions include the best tolerated dosing ranges; the manner in which the drug is absorbed, distributed, metabolized, and excreted; as well as the duration of the drug's therapeutic action. During Phase II, investigators use a battery of tests among 200 to 300 patients to obtain convincing evidence of the drug's medical benefits. Controlled tests are often used to measure the drug's effects against a placebo and are designed as open-label, or blinded, studies. Blinded studies are used to reduce subjective bias during analysis of a new compound. The total development time to this point averages six and one-half years. Once again, many compounds are dropped.
Drugs that survive the first two phases of human testing enter Phase III
clinical testing to reconfirm whether the drug is effective and to identify any side effects that occur in statistically significant numbers of patients. Phase III studies are extensive and involve 1,000 to 3,000 volunteer patients in clinics and hospitals nationwide. The average length of time for Phase III testing is three years, which brings the time of development thus far to nine and one-half years. Products still fail and are terminated at this point.
REDUCING EXPOSURE OF PRODUCTS IN DEVELOPMENT
Measures to reduce exposure to liability are built into each research stage. Clinical investigators are indemnified. Institutional Review Boards are established at investigational sites to implement programs that monitor patient safety and ensure patient rights. Patients who volunteer for clinical trials are protected by stiff federal regulations and ethical standards. Federal regulation requires patients to be fully informed of the aims, methods, anticipated benefits, and potential hazards of a study before enrolling. Investigators must ensure that volunteers understand they are free to refuse to enter a study or to withdraw at any time. Written consent must be obtained from patients before they may participate. In the case of diminished ability, written consent is obtained from the patient's relatives or legal guardian. These measures appear to be effective: notwithstanding the recent experience with clinical trials of fialuridine as a treatment for hepatitis B, a review of case law suggests that manufacturers have not faced substantial litigation by clinical trial participants (Reisman, 1992).
Upon completion of Phase III trials, the developer files a New Drug Application (NDA) with the FDA to obtain a license to market the product for general use. In traditional paper form, the NDA can reach 90,000 pages and may actually fill an entire truck. In addition to providing the results of clinical testing, the NDA must include the suggested product labeling and drafts of advertising and promotional materials for FDA approval. While the FDA is required by law to review the NDA within six months, the average time for this process is two to three years.
THE HIGH PRICE OF REGULATION
This extended discussion of the drug development process is intended to prove a point: the safety and efficacy of a new drug have been rigorously evaluated prior to its entry into the market. Its medical benefits in relation to its potential risks have been calculated, weighed, and sanctioned. Clear, consistent product labeling has been developed to educate prescribers and encourage responsible administration of new drugs, thereby
further reducing liability exposure. The developer has earned a reasonable assurance of protection against liability.
But at what price? Twelve years have passed. The pharmaceutical developer has spent $359 million without realizing a dime. And, liability is most likely to occur when an approved drug is used in populations that have not been substantially studied (Levine, 1993).
This is only one reason why clearance for marketing does not mark the end of the regulatory process. Post-marketing surveillance, or Phase IV, studies may take place to study emergence of new side effects or use of the drug in patient groups not previously explored. Additional studies may be conducted to compare a new drug with existing medications. Pharmaceutical developers may also wish to pursue new medical applications, new claims, or formulations of the drug beyond its approved indication. These studies may be conducted under the original, or new, IND as additional rigorously controlled Phase III trials, and may take up to four years. The high cost of additional trials often discourages manufacturers from seeking expanded indications.
Not even explored in this paper are the indirect costs to society of the regulatory process, including global competitiveness, the impact of the regulatory process on drug pricing, the limited access of critically ill patients to therapies in development, and reimbursement issues that arise when physicians prescribe drugs for off-label applications. Also not addressed is the disincentive to innovation caused by the fact that the regulatory process erodes patent protection. Another particularly critical issue is the need to ensure incentives for development of orphan drugs, or medicines for rare diseases with small patient populations.
IS REGULATION WORTH THE PRICE?
Are the enormous costs and painstaking pace of current regulation worth it? Does the grueling process pay off in reduced liability exposure for pharmaceutical products?
There is a widespread perception that the pharmaceutical industry is the victim of a nationwide litigation explosion. Indeed, a recent RAND study found that pharmaceuticals was a leading industry in federal liability suits, with a strong surge in case filings during the 1980s (Dungworth, 1988). But close analysis of the filings shows 60 percent of cases involved only two products, the Dalkon Shield and Bendectin. Further analysis shows that the Dalkon Shield and Bendectin rank second and third only to asbestos in terms of growth of federal product liability filings from 1974 to 1985 (General Accounting Office, 1988). These data suggest that instead of a litigation explosion, the pharmaceutical industry is vulnerable to concentrated clusters, or epidemics, of litigation. Cases involving Prozac,
used to treat depression, and Halcion, for the treatment of insomnia, are current examples of this phenomenon.
CHILLING LESSONS OF BENDECTIN
The value of the tort liability system in situations where deception, fraud, or latent injury emerge is unquestioned. But the case of Bendectin is a cautionary tale.
Bendectin was the only prescription drug ever approved in the United States for the treatment of nausea and vomiting in pregnancy. Introduced in 1956, Bendectin was used in more than 30 million pregnancies. Beginning in 1969, assertions that Bendectin could produce congenital birth defects began to appear in scientific literature. While no sound scientific study ever proved a causal relationship between Bendectin and birth defects, and the FDA continued to affirm its safety, nearly 1,700 lawsuits were brought against the manufacturer (Pharmaceutical Manufacturers Association, 1993). The manufacturer won almost every case that went to court, but the price was too high. In 1983 the manufacturer voluntarily withdrew Bendectin from the marketplace because its $18 million annual cost for legal fees and insurance began to overwhelm its $20 million in annual sales (Viscusi and Moore, 1991). It is unlikely that any new drug will be developed to close this therapeutic gap.
DISPARITIES BETWEEN GOALS OF REGULATION AND LITIGATION
The Bendectin case illustrates disparities between the goals of the regulatory process and the tort liability system. The tort liability system is designed to discover risk and assign blame to a probable cause of injury after it has happened. On the other hand, the regulatory system seeks to predetermine potential adverse effects of a pharmaceutical product in order to prevent or manage their occurrence. In the case of Bendectin, assertions that it was a teratogen, or agent that causes birth defects, were based on the argument that if a pregnant women took Bendectin and if she gave birth to a deformed child, it was possible that Bendectin was the cause. This argument was designed to divorce each case from the backdrop of epidemiological data showing that chance alone could account for the incidence of 900,000 births of malformed babies among the 30 million women who took Bendectin while pregnant. It is interesting to note that the rate of birth defects has not declined in the United States since Bendectin was withdrawn. However, treatment for severe nausea during pregnancy now accounts for nearly $40 million of the nation's annual hospital bill (Pharmaceutical Manufacturers Association, 1993).
AREAS FOR TORT REFORM
Reforms are needed in many areas. One in particular—the issue of expert testimony—is indicative of the previously mentioned asymmetry between the objectives of regulation and those of product liability law. Courts have allowed expert witnesses to offer subjective or anecdotal testimony that is not based on sound, peer-reviewed scientific data. This practice of "junk science" has led to high jury awards in cases where no scientific evidence substantiated fault on the part of the manufacturer. The debate over "science in the courtroom" led to the 1993 Supreme Court ruling that federal judges must ensure that scientific evidence and testimony admitted in trials is "not only relevant, but reliable … that the methodology underlying the testimony is scientifically valid … that the theory or technique has been subjected to peer review" and that the ''known or potential rate of error" of a particular scientific technique is considered.4
While the dismissal of "junk science" from federal courtrooms may eventually be good news for the industry, reform is needed in other areas of pharmaceutical product liability as well. The Pharmaceutical Manufacturers Association calls for the creation of a uniform federal tort liability system, as opposed to the current patchwork of state laws and the barring of punitive damages against manufacturers if they have already met the stringent requirements of the FDA. Other nations have diminished the lure of the industry's deep pockets and the impact of high jury awards by instituting social insurance schemes to redress injury. The national vaccine injury compensation program, which went into effect in the United States in 1988, resembles such measures and may signal the beginning of a trend.
A VIABLE EQUATION
It is true that the U.S. regulatory process and tort liability system greatly affect the cost of drug development and the length of time to market. The only real window for innovation is at the beginning of the 12-year regulatory process. The ultimate benefit of this process, however, is assurance of reasonably safe, effective prescription drugs.
Enormous strides in medical science have occurred in the past few decades despite the constraints of the system. Diseases have been eradicated. The prognosis for patients with diseases such as cancer and heart disease has dramatically improved. Biotechnology ushers in a new era of exciting possibilities, including the potential for safer, specifically targeted vaccines. The health of the industry appears to indicate that regulation, rational
tort liability, and innovation represent a viable equation instead of an impossible triangle. Ongoing medical progress is telling evidence of a system that works.
Dungworth, T. 1988. Product Liability and the Business Sector: Litigation Trends in Federal Court. R-3668-ICJ. Santa Monica, Calif: RAND.
Fern, F. H., and W. M. Sichel. 1985. Failure to warn in drug cases: are punitive damages justifiable?" For the Defense 27:12–20.
General Accounting Office. 1988. Product liability; extent of "litigation explosion" questioned. Publication no. GAO/HRD-88-36 BR. Washington, D.C.: U.S. Government Printing Office.
Lasagna, L. 1991. The chilling effect of product liability on new drug development. Pages 334–359 in The Liability Maze: Impact of Liability Law on Safety and Innovation, P. W. Huber and R. E. Litan, eds. Washington, DC.: Brookings Institution.
Levine, C. 1993. Women as research subjects: New priorities, new questions. Pages 169–188 in Emerging Issues in Biomedical Policy: An Annual Review, R. H. Blank and A. L. Bonnickson, eds. Vol. 2. New York: Columbia University Press.
Pharmaceutical Manufacturers Association. 1991. Industry Issues Brief: The Drug Approval Process. Washington, D.C.: Pharmaceutical Manufacturers Association.
Pharmaceutical Manufacturers Association. 1993. Industry Issues Brief: Product Liability Reform. Tort Reform File. Washington, D.C.: Pharmaceutical Manufacturers Association.
Reisman, B. K. 1992. Products liability—What is the current situation and will it change (and how) when more women are included in studies? Presented at the Women in Clinical Trials Workshop of the FDA-Regulated Products Workshop, Food and Drug Law Institute, Washington, D.C.
Swazey, J. P. 1991. Prescription drug safety. Pages 291–333 in The Liability Maze: Impact of Liability Law on Safety and Innovation, P. W. Huber and R. E. Litan, eds. Washington, D.C.: Brookings Institution.
U.S. Congress, Office of Technology Assessment. 1993. Pharmaceutical R&D: Costs, risks and rewards. Publication no. OTA-H-522. Washington, D.C.: U.S. Government Printing Office.
Viscusi, W. K., and M. J. Moore. 1991. Rationalizing the relationship between product liability and innovation. Pages 105–126 in Tort Law and the Public Interest, P. H. Schuck, ed. New York: American Assembly, Columbia University.