1
Introduction
The definition of insomnia remains somewhat unsettled both in clinical practice and in research. Several definitions currently exist, including those in the International Classification of Sleep Disorders. Insomnia is usually defined as a subjective complaint of poor, insufficient, or nonrestorative sleep. The duration of symptoms is important in the evaluation of a complaint of insomnia. Transient or short-term insomnia is common in otherwise healthy people who have acute stress; who are bereaved, jet lagged, or admitted to a hospital; or who have an intercurrent illness. Long-term insomnia, usually defined as at least 2 or 3 weeks in duration, is often associated with chronic medical conditions. Chronic insomnia is often multifaceted and has multiple determinants. In the clinical evaluation of the insomniac patient, clinicians often try to identify predisposing, precipitating, and perpetuating factors.
Insomnia is a common symptom of many disorders; about one third of the adult population experiences insomnia each year. By most recent estimates, there appears to be approximately a 10 percent prevalence of chronic insomnia in adults in the United States (Mellinger et al., 1985). It is also estimated that the annual cost of treatment of insomnia is between $92.5 billion and $107 billion in the United States (Staller, 1994). Reports of insomnia tend to increase with age and are more prevalent among women. Furthermore, people who are divorced, widowed, or separated have chronic insomnia more often than married individuals. Finally, insomnia is persistent and recurrent in both clinical populations and community-based samples of the population (Balter and Uhlenhuth, 1992).
Given the need to treat chronic insomnia, the treatment approach has generally been use of sedative-hypnotic medication. It is estimated that such forms of medication are used by 5 percent of the population in a year's time and over the course of a year are used regularly by 0.5 percent of the population in the United States (Mellinger et al., 1985; Baltex and Uhlenhuth, 1992). Despite the widespread use of pharmacotherapy in the treatment of insomnia, the optimal role of medication is still poorly defined (Bliwise, 1991). Attempts to provide educational and behavioral interventions are receiving increasing attention from the medical
community. Various types of educational efforts and behavioral intervention, including stimulus control therapy, sleep restriction therapy, and improved sleep hygiene, have been shown to provide considerable benefit (Kupfer and Reynolds, 1997).
HISTORICAL OVERVIEW OF HALCION
Upjohn,1 a company that manufactures pharmaceutical products including hypnotic drags, submitted a New Drug Application (NDA) to the U.S. Food and Drug Administration (FDA) for its sleep-inducing drug, triazolam, in May 1976. Subsequently marketed under the trade name Halcion,2 triazolam was a new compound in the benzodiazepine group—a category of sedative-hypnotic agents (see Table 1-1). Benzodiazepines represented a great improvement over barbiturates in terms of both efficacy and safety. Additionally, triazolam appeared to be an improvement over other benzodiazepines because of its short half-life. Because the body eliminated it rapidly, triazolam did not induce the hangover effect associated with other benzodiazepines. (Box 1-1 further describes the pharmacology of triazolam.) As a consequence of these advantages, physicians prescribed Halcion widely, and the drug became one of Upjohn's best-selling pharmaceuticals (U.S. Food and Drug Administration, 1996).
TABLE 1-1 Common Benzodiazepines and Their Trade Names
Compound |
Trade Name |
alprazolam |
Xanax |
diazepam |
Valium |
flurazepam |
Dalmane |
lorazepam |
Ativan |
midazolam |
Versed |
oxazepam |
Serepax |
zolpidem |
Ambien |
temazepam |
Restoril |
triazolam |
Halcion |
Before Halcion reached the U.S. market, however, it had undergone an eventful approval process (see Box 1-2 for a detailed time line). Upon reviewing the NDA, the chief medical review officer expressed concern about both efficacy and safety, including (1) high rates of amnesia, incoordination, confusion, and other central nervous system-related side effects associated with Halcion (0.5 to 1.0 mg) compared with those associated with either
flurazepam or placebo; (2) possible diminished effectiveness with repeated administration; (3) "rebound insomnia" during withdrawal; and (4) possible unique safety or efficacy profiles uncharacteristic of those for other benzodiazepines. These concerns were not sufficient, however, to prevent an FDA Psychopharmacologic Agents Advisory Committee (now named the Psychopharmacologic Drugs Advisory Committee [PDAC]) from recommending in 1977 that Halcion be approved for clinical use. The approval was delayed, however, by a request from the Division of Neuropharmacological Drug Products for additional preclinical animal studies and human bioavailability data. When FDA approved Halcion for clinical use in November 1982, reports of possible problems with the drug were already starting to appear in Europe.
BOX 1-1 PHARMACOLOGY OF TRIAZOLAM Triazolam (Halcion) is known for its rapid onset of action when given orally, for its short duration of action, and for its ability to produce a sleeplike state (sedation). This profile makes it a candidate for effective therapeutic use in patients with sleep disorders. Other benzodiazepines that have been used therapeutically include chlordiazepoxide, clonazepam, clorazepate, diazepam, flurazepam, lorazepam, quazepam, and temazepam. The profile of activity of the benzodiazepines includes muscle relaxation, sedation, anxiolytic activity, anticonvulsant activity, and induction of amnesia. All types of activities can be demonstrated among the benzodiazepines, with one or more attributes dominating and being characteristic for each compound. Each drug is further characterized by the time of onset of activity and the duration of the effects. The duration of action of each drug is characterized by its rate of metabolism and clearance from the body, by the possible generation of active metabolites, and by excretory mechanisms. Importantly, triazolam is metabolized by cytochrome P-450 3A4, an enzyme with activity that is both inducible and inhibitable by a large number of other drugs and even grapefruit juice. |
Overseas Events
Halcion was approved for use at a dose of 1.0 mg in The Netherlands in 1977. Two years later, a Dutch psychiatrist, C. van der Kroef, published an article (1979) detailing a series of adverse events in his patients that he described as a syndrome. The syndrome included symptoms of depression, amnesia, hallucinations, and anxiety. Although van der Kroef never disclosed his source records, The Netherlands' regulatory body suspended Halcion from its market and began negotiations with Upjohn to change the labeling, restrict its use, and require further studies. In February 1980, Upjohn withdrew Halcion from the Dutch market rather than comply with the possible restrictions.
In the United Kingdom, a decision to review the safety of Halcion was made in response to a report from Upjohn that "errors had been identified in a report of one of the clinical studies included in the original" application for approval (U.S. Food and Drug Administration, 1996, p. 1). The summary tables for a tolerability study with prisoners (Study 321) did not reflect all of the adverse events reported by the participants. Upon learning this,
many governments around the world reassessed the safety of the drug, and most made changes in the labeling. This included the Committee on Proprietary Medicinal Products (CPMP) of the European Union. CPMP released two position papers (in October and December 1991) that dealt specifically with Halcion. In both papers, CPMP concluded that the drug was safe as labeled, but warned that the use of Halcion should precisely follow the labeling in terms of the maximum dose (0.25 mg) and the length of use (not more than 10 days). Additionally, CPMP formed an ad hoc group in 1993 to assess Halcion, along with six other short-acting hypnotic drugs, at the request of France and The Netherlands. This group saw a risk of dependence on these drugs and felt that these drugs should be used only for short periods of time and only for severe circumstances. The use of Halcion, specifically, should be limited to a few days with a maximum of 2 weeks. The United Kingdom took the additional step of suspending Upjohn licenses to market the drug and, after further review, revoked them permanently in 1993.
FDA Activity
FDA approved Halcion for use in 1982 and since then has monitored its safety through postmarketing surveillance, a large part of which was through the Spontaneous Reporting System (SRS), the system that FDA uses to track adverse events reported by physicians and patients. 3 FDA also reassessed the data from the NDA, and, in 1994, formed a task force to examine various related issues.
Halcion had been associated with a large number of these spontaneous reports of adverse events from the time of its approval. Indeed, there was sufficient concern about the safety of the drug to cause Public Citizen, a consumer advocacy group, to submit petitions in 1991 and 1992 requesting that FDA remove Halcion from the U.S. market. 4
Spontaneous Reports
Interpretation of the spontaneous reports was a matter of some debate within FDA. The FDA's Office of Epidemiology and Biostatistics identified a signal from the SRS data; that is, there appeared to be a heightened number of reports of adverse events associated with Halcion within the first six years of the drug's marketing. For example, in that period, there were 174 spontaneous reports of amnesia by people taking Halcion whereas there were only 3 for temazepam (Restoril), another benzodiazepine hypnotic drug. Even though there were approximately 10 million more prescriptions for Halcion in the compared time frames, the data suggested an increased rate of adverse events for Halcion (Tsong, 1992). In addition, Wysowski and Barash (1991) reported that psychiatric adverse reactions were up to 56 times higher (amnesia) with Halcion than with temazepam (see Chapter 3, Table 3-14).
The FDA Division of Neuropharmacological Drug Products, however, questioned the reliability and interpretation of the SRS data, suggesting that the large number of reports reflected a general tendency to overreport adverse events, particularly following adverse media attention, rather than a greater incidence in the actual occurrence of adverse events. An analysis performed by Tsong (1992) used statistical methods to adjust for many factors, including publicity, variability in rate ratios, manufacturers' reporting practices, and trends in overall reporting rates, among others. This analysis resulted in reduced overall risk ratios (e.g., 8 to 1 relative risk for depression and 26 to 1 relative risk for amnesia). Note, however, the rate ratio for seizures, an unlikely adverse event for a benzodiazepine, was as high as 26 to 1, and mortality was 3 to 1. In 1992, PDAC assessed these data and concluded that action was not required on the part of FDA.
FDA Reassessment
In addition to tracking the postmarketing data, FDA reassessed the information from the original application, which Upjohn had reentered into a new database, and requested that Upjohn undertake an integrated summary of safety. FDA also investigated Upjohn "for adherence to applicable laws and regulations in generating clinical safety and efficacy data for Halcion," an effort that was subsequently handed over to the U.S. Department of Justice. Finally, PDAC reviewed the data in May 1992. That committee judged the numbers from the new database to be valid and Halcion to be safe and effective, but at a lower dose: 0.25 mg for the general population and 0.125 mg for the elderly population. In response to a request by PDAC, Upjohn initiated two new dose-response studies (Protocols M/2100/0366 and M/2100/0373) and one large insomnia treatment study (Protocol M/2100/0235) comparing the safety and efficacy of Halcion compared with those of temazepam.
FDA Task Force
In 1994 FDA formed a task force to investigate scientific questions, regulatory concerns, and possible misconduct or impropriety by both Upjohn and FDA in the Halcion approval process. The FDA task force made recommendations related to improving the drug approval process and improving the investigation of suspected misconduct. It concluded that Upjohn's actions should have been considered material5—but that this in and of itself does not constitute a violation of the law.
With respect to safety and efficacy, the task force concluded that Halcion was "safe when prescribed according to current labeling" and "effective in the treatment of insomnia at doses and durations currently recommended in the labeling" (U.S. Food and Drug
Administration, 1996, p. iii). The task force also recommended that a separate reassessment of the safety and efficacy of Halcion be conducted by a panel of independent experts. To that end FDA requested the present study by the Institute of Medicine (IOM).
CHARGE TO THE IOM COMMITTEE
FDA asked IOM to perform an independent assessment of the publicly available data on Halcion and to prepare a report. This was to include an assessment of the following:
-
the adequacy of the study designs and quantitative endpoints used in the major clinical trials of Halcion;
-
the quality and quantity of postmarketing data with respect to adverse drug reactions;
-
the overall confidence in the data on the effectiveness, adverse events, and side effects of Halcion at different doses and for different durations, including those specified in the current product labeling; and
-
the need for additional studies to clarify and characterize the risk and efficacy profiles of Halcion.
If additional studies were deemed necessary to help clarify and characterize the risk and efficacy profiles of Halcion, the committee was instructed to describe what specific types of studies would be needed.
ORGANIZATION OF THE REPORT
The remainder of this report is organized into sections that address issues related to the quality and adequacy of the data regarding effigy (Chapter 2) and safety (Chapter 3). The final chapter (Chapter 4) presents concluding remarks regarding broader implications. Several appendixes (A-G) are included, as follows: Appendix A, FDA safety tables; Appendix B, summary tables of literature reviewed for safety of Halcion; Appendix C, glossary; Appendix D, acronyms; Appendix E, resources reviewed by the committee; Appendix F, Upjohn consent to disclosure; and Appendix G, committee and staff biographies.
BOX 1-2 TIME LINE OF SIGNIFICANT EVENTS IN HALCION'S HISTORY
|
|
|
|