Seven pillars of a new evidentiary paradigm: the Food, Drug, and Cosmetic Act enters the genomic era.

• Author: Evans, Barbara J.

To assess the impact of the March 2009 decision in Wyeth v. Levine, it is crucial to understand that the Supreme Court ruled on actions that the U.S. Food and Drug Administration (FDA) took under a statutory scheme that already had been amended by the time the case was decided. The Food and Drug Administration Amendments Act of 2007 (FDAAA) transformed drug regulation, adding significant new powers to develop evidence and make new types of decisions in the postmarket period. This article explores how the contours of drug regulation are likely to change after FDAAA, which is the most profound reworking of the U.S. drug regulatory framework in half a century.

FDAAA envisions heavy use, during the period after drugs are approved, of evidence from large observational studies that rely on interoperable health data networks. Understanding what was wrong with FDA's old evidentiary paradigm, which dates back to 1962, is essential to understanding its new one. Parts II and III of this article discuss the evidentiary limitations of premarket drug trials; important aspects of modern legal doctrine rest on misconceptions about their evidentiary power. Part IV then explores how scientific advances flowing from the Human Genome Project over the past decade further undermined FDA's old evidentiary paradigm. FDAAA was Congress's response to these problems. Part V identifies seven pillars of the new evidentiary paradigm: seven novel propositions that reject foundational assumptions of twentieth-century drug regulation. Collapse of these assumptions sets off ripple effects in various doctrinal areas. Part VI provides two examples, with the aim of opening a scholarly debate about these and other impacts of FDA's new evidentiary paradigm.

INTRODUCTION I. THE 1962 EVIDENTIARY PARADIGM FOR DRUG REGULATION A. The Gap Between Perception and Performance B. The Role of Randomized, Controlled Trials C. Evidentiary Paths Not Taken in 1962 II. LIMITATIONS OF THE OLD EVIDENTIARY PARADIGM A. The Myth of the Premarket Safety Trial B. Methodological Limitations of Premarket Safety Studies III. THE OLD PARADIGM: REPAIR OR REPLACE? IV. THE CHALLENGES OF GENOMIC MEDICINE A. The Challenge of Predictive and Preventive Medical Technologies B. The Challenge of Heterogeneous Drug Response V. SEVEN PILLARS OF THE NEW EVIDENTIARY PARADIGM A. Postmarket Evidence Development B. The Flexible Best Evidence C. Successive Improvement of the Risk-Benefit Ratio D. Putting Evidence into the Hands of Innovators E. Making Efficacy Regulation Effective F. Piercing the Veil of Average Safety and Efficacy G. Making Evidence Consequential VI. UNRESOLVED DOCTRINAL ISSUES A. Preemption of Suits Against Manufacturers After FDAAA and Wyeth v. Levine B. Federal Intrusion on Medical Practice CONCLUSION INTRODUCTION

The Food and Drug Administration Amendments Act of 2007 (1) (FDAAA) transforms the evidentiary basis of medical product regulation by the U.S. Food and Drug Administration (FDA). FDAAA augments premarket clinical studies with new sources of evidence about the risks and benefits of drugs. FDAAA envisions heavy use, during the postmarket period, of large observational studies that rely on interoperable health data networks. This shift will affect diverse areas of legal doctrine.

Product liability and medical malpractice are obvious candidates for impact. Questions about the preemptive effect of FDA's drug regulation loom large after FDAAA, notwithstanding the U.S. Supreme Court's March 2009 decision in Wyeth v. Levine. (2) "The major doctrinal question here is whether various forms of regulatory action by the FDA have the effect of preempting suits brought pursuant to state tort law." (3) Wyeth v. Levine held that FDA's drug approval and labeling decisions did not preempt a failure-to-warn suit against the drug's manufacturer. That result seems fair, given that the evidence on which FDA has been basing these decisions is severely limited in ways this article explores. (4) Congress enacted FDAAA to address these very limitations. FDAAA expands FDA's postmarket evidence development and empowers FDA to take new forms of regulatory action in response to the new flows of evidence. Some of these new regulatory actions may have the preemptive effect that the Court, in Wyeth v. Levine, found lacking in FDA's pre-FDAAA approval and labeling decisions. In assessing the impact of Wyeth v. Levine, it is crucial to understand that the Supreme Court was ruling on actions taken under a statutory scheme that had been substantially amended by the time the case was decided.

FDAAA directs FDA to address certain matters that Congress previously regarded as state medical practice issues. There may be constitutional questions as FDA flexes new powers to restrict clinical use of drugs and does so in ways that deny seriously ill patients access to FDA-approved pharmaceutical products--a problem that appears already to be occurring. FDA-imposed use restrictions raise the specter that physicians, as well as manufacturers, might argue preemption--or its close cousin, the regulatory compliance defense (5)--in drug-injury malpractice suits. In other areas, FDAAA breathes new life into classic infrastructure regulatory problems by requiring evidence that can only be generated with a massive, networked informational infrastructure that does not yet exist and will have to be financed, built, and administered. (6) The new forms of evidence, as they become available, may reshape cultural and commercial norms that have favored avid consumption of prescription drugs (7) in recent decades, and they may alter popular expectations of medical privacy. (8) The new research methodologies may require new structures for research oversight and new principles for the ethical conduct of biomedical research. This article touches on these broader impacts to open a scholarly dialogue about them. However, it focuses on how the broad, conceptual contours of drug regulation have changed and why. Understanding what was wrong with FDA's old evidentiary paradigm is key to appreciating the promise and the perils of its new one.

FDAAA is a profound change in law. The Federal Food, Drug, and Cosmetic Act (9) ("FDCA" or "1938 Act") has been amended more than one hundred times (10) since it became law in 1938. Most of these amendments were "tweaks" but, occasionally, Congress fundamentally altered the regulatory paradigm. FDAAA was a "sweeping overhaul of both the FDCA and the Public Health Service Act." (11) Once called "the biggest reforms since at least 1997," (12) FDAAA is in fact the most momentous shift in drug regulation in half a century. The 1962 Drug Amendments (13) sparked the last big shift by introducing the modern concept of drug approval. Before 1962, new drugs were deemed approved for sale unless FDA, within a limited period of time, found grounds to keep them off the market. (14) Since 1962, drugs await affirmative approval before they can be sold. (15) It was after 1962 that "FDA became responsible for making benefit-risk decisions" about drugs. (16) Congress directed FDA, for the first time, to require "substantial evidence" (17) of efficacy as well as safety. (18) FDA interpreted (19) this language to require the familiar three-phase clinical trial process (20) through which drugs now pass before FDA approval.

These requirements contributed to expansive growth of clinical trial activity late in the twentieth century. Between 1962 and 1980, the cost to develop a new drug rose from $6.5 million to $70 million, (21) with much of the increase attributed to the cost of FDA's required clinical studies. (22) In 2007, the Institute of Medicine reported various estimates ranging from $500 million to $2 billion as the cost of developing one new drug. (23) These data suggest a clear upward trend although historical and recent figures are not strictly comparable. (24) Robert Califf traces how FDA's clinical trial requirements led to a "complex maze of rules ... and ultimately created an entirely new industry: the contract research organization (CRO)" (25) to help drug sponsors navigate FDA's premarket requirements. As of March 2008, the CRO industry had estimated annual revenues of $17.8 billion, employed 100,000 people worldwide, and was experiencing annual revenue growth of 14-16%. (26)

In their day, the 1962 amendments marked the most significant shift in regulatory philosophy not only since the 1938 Act, but since the 1906 Pure Food and Drug Act (27) that had preceded it. (28) Another important shift was the 1976 Medical Device Amendments (29) which authorized FDA to regulate medical devices, a category that later has acquired significance in genomic medicine since it includes diagnostic and genetic tests. FDAAA belongs in this class of paradigm-shifting amendments.

If the 1962 amendments heralded the age of faith in clinical trials, then FDAAA tolls its twilight. That, in a nutshell, is the shift of regulatory philosophy now underway. It is important to qualify what that statement means: Declaring the twilight of premarket clinical trials is not the same thing as saying that the sun already has set, or will ever set, on them. Clinical trials will still be performed and, indeed, FDAAA expanded FDA's authority to require them in the postmarket as well as the premarket period. (30) Clinical trials will continue to be an important--often, the primary--source of data about medical products. Yet FDAAA accepts that clinical trials have intrinsic limitations and cannot, by themselves, ensure the safety and efficacy of FDA-approved products. FDAAA is a response to this reality. Premarket trials will continue as before, but there has been a pragmatic reassessment of their evidentiary value. One might refer to this as the nation's Trialdammerung--its loss of faith in evidence from premarket drug trials. Derived from "clinical trial" and the German word, Dammerung (meaning "dimming" or "twilight" (31))...