An examination of the published information relating to the use of drugs in the treatment of obesity reveals several of the underlying realities for treatment of this disease [1]. First, obesity is a chronic disease. A steady, slow, progressive increase in body weight occurs during adult life. In addition, obesity has many causes. Although most patients do not have an easily ascribable cause for their obesity, clear cases exist in which obesity is caused by genetic alterations, neurologic disorders, endocrine disease, or drugs. Because of its chronic nature and unknown cause, obesity is rarely cured; however, palliation is a realistic goal. Weight loss occurs slowly except when the most drastic means are used. Recidivism, that is, weight regain, is common after termination of a weight-loss program. Compared with the relatively slow rate of weight loss, weight gain may be more rapid. Drugs have been used for the treatment of obesity, but regain after cessation of treatment is often ascribed to failure of the drugs. A more appropriate conclusion is that medications do not work if not taken. This principle applies to medications used to treat obesity as it does to medications used to treat hypertension, diabetes, heart disease, asthma, or gastrointestinal illnesses.

As one searches for agents that modify body fat, the most exciting possibilities lie in the agents that alter nutrient partitioning (that is, those that partition nutrient intake into protein formation with a reduction of body fat). Such agents might include growth hormone, which enhances milk production and growth. Anabolic steroids also have the potential for enhancing protein formation relative to fat, as do some of the ß-2 and ß-3 adrenergic agonists, which have been shown in animal models to increase lean tissue relative to body fat. These effects can be likened to the effects of long-term exercise, which enhances fat oxidation, reduces body fat stores, and increases muscle mass. Agents with these properties would be ideal for the treatment of obesity.

Historical Perspective

The medical history of obesity dates back to the Stone Age [2]. The primary approaches to treatment from the time of Hippocrates until the 20th century have involved changes in diet and exercise. Before pharmacologic treatment, the most famous diet was that published by Banting in 1863 [3]. During the 20th century, many new treatments for obesity, including drugs, have appeared [4]. The first of these, thyroid hormone, was introduced in 1893 and was believed to be therapeutic because many overweight patients were thought to have a low metabolic rate. This treatment was widely used until after World War II. Two other treatments were also introduced before World War II. The first of these was dinitrophenol, a drug that was noted to increase metabolism and to produce weight loss. The mechanism for this effect was shown to be the uncoupling of oxidative phosphorylation. This drug was abandoned because of severe side effects including neuropathy and cataracts. Dextroamphetamine, introduced in the 1930s, was originally developed as a treatment for narcolepsy, but by serendipity was noted to produce weight loss. It was subsequently found to produce weight loss by reducing appetite, thus initiating a new era of pharmaceutical therapy for obesity [4].

Appetite-suppressing Drugs

Noradrenergic Drugs

Amphetamine, the first noradrenergic agent, has many of the pharmacologic properties of the naturally occurring transmitters—norepinephrine, epinephrine, and dopamine. Amphetamine, however, suffers from the serious disadvantage of having a substantial abuse potential [5]. Abuse of amphetamine, methamphetamine, and (to a lesser extent) phenmetrazine has given this entire group of drugs a bad name. No clinical need exists for any of these drugs in the treatment of obesity. The central excitatory properties of amphetamines led to the search for derivatives containing the appetite-suppressant effects without the potential for drug abuse. Chemical manipulations of the side chain and ring structure of amphetamine have markedly reduced the risks for central nervous system stimulation and abuse but have retained the appetite-suppressing effects. Because these drugs mimic norepinephrine, they are listed as noradrenergic drugs in Table 2. All these drugs have a similar spectrum of pharmacologic action.

In a review of the effectiveness of amphetamine-like drugs in the early 1970s, the Food and Drug Administration reanalyzed all controlled studies claiming effectiveness for an anoretic drug submitted to that agency [6]. More than 200 such studies including almost 10 000 patients in trials lasting 4 to 20 weeks were identified. This project included an analysis of 105 new drug applications. In more than 90% of the studies, the active drug produced more weight loss than did placebo, but the drug-related weight loss was neither great nor invariable. In 40% of the 160 trials comparing placebo with active drug, the patients receiving active drug lost significantly more weight than did those receiving placebo. A pool of all patients included in the parallel studies yielded 4543 who were treated with active drug and 3182 who were given placebo (Table 3). After 4 weeks of therapy, the dropout rate for patients receiving active drug was 24.3% compared with 18.5% for those receiving placebo. Forty-four percent of patients receiving active drug lost 0.45 kg (1 lb) or more per week compared with 26% of patients receiving placebo; 2% receiving active drug lost more than 1.35 kg (3 lb) per week compared with only 1% of those receiving placebo. The average weight loss for patients taking the drug was 0.23 kg (0.5 lb) per week greater than that for those receiving placebo after 4 weeks of therapy. After combining data for all patients at their final weights, Scoville [6] found that patients taking active drug lost approximately 0.23 kg (0.5 lb) per week more than did those taking placebo. He noted that "because drugs do not provide complete cures, however, is no reason to reject them out of hand; partial success is clearly better than failure."

Table 4 summarizes the short- and long-term published studies of drugs used in the treatment of obesity [7–66]. The table includes drugs scheduled by the Drug Enforcement Agency of the U.S. government as well as nonscheduled drugs that are available over the counter or that have not yet been approved for use in the treatment of obesity. In all but three studies, the drug-treated group lost more weight than did those receiving placebo. Several studies reported weight losses of 8 kg or more above that of the placebo group; however, the differences in most studies were smaller [67].

Serotonergic Drugs

The introduction of fenfluramine, a drug that resembles the amphetamines chemically but not pharmacologically, opened a new chapter in pharmacologic therapy of obesity. Fenfluramine acts to partially inhibit the reuptake of serotonin and to release serotonin from nerve endings. This increased serotonin in the neuronal cleft is believed to reduce food intake. Several currently existing mechanisms will enhance excess neuronal serotonin concentration and also inhibit food intake [68]. The mechanisms of action are diverse and range from the precursors of serotonin action to drugs that primarily inhibit serotonin reuptake into nerve terminals, drugs that have a substantial effect on serotonin release as well as on serotonin uptake, and drugs that act directly on serotonin receptors. Of these drugs, d,l-fenfluramine and D-fenfluramine are available for clinical use in the United States and Europe, respectively. Several long-term studies with fenfluramine have been reported [18, 23, 26, 44]. Among patients who remain in the trials, fenfluramine is clearly more effective than placebo in producing and maintaining weight loss for periods up to 1 year. In a double-blind, randomized trial, fenfluramine combined with phentermine enhanced the magnitude of weight loss over and above what could be achieved with the best behavior modification, exercise, and nutrition program, and the effect continued for nearly 4 years [64]. The principal drawback to the use of fenfluramine is the development of depression when the drug is withdrawn abruptly; rare reports of pulmonary hypertension have also been documented [69-72].

Fluoxetine is an inhibitor of serotonin uptake that is currently marketed as an antidepressant. A second serotonin reuptake inhibitor, sertraline, has recently been approved as an antidepressant and is reported to produce weight loss. During trials to determine the effectiveness of fluoxetine as an antidepressant, fluoxetine-treated patients were noted to lose small amounts of weight. Subsequently, a number of clinical trials have been published in which the drug was used to treat obesity [15, 21, 24, 35, 36, 47]. Fluoxetine shows a dose-response effect on weight loss over 8 weeks of treatment [35]. When patients were treated with the drug for more than 16 to 20 weeks, however, most began to regain weight [15, 24]. Although the mechanism for this weight regain is unclear, it does not occur in all patients; therefore, fluoxetine may be useful for continued therapy in the subset of patients who do not regain weight.

Newer Pharmacologic Agents

Our understanding of the mechanisms of obesity has advanced rapidly in the 40 years since the end of World War II. Based on this understanding, it is now clear that one can manipulate body fat in several ways [73, 74]. Currently available drugs reduce food intake through adrenergic or serotonergic mechanisms. The long-term effectiveness of the noradrenergic drugs has not been adequately evaluated. The combined use of a serotonergic drug, d,l-fenfluramine, and a noradrenergic drug has been shown to cause weight loss for more than 3 years [64]. The serotonergic drug dexfenfluramine has been shown to work for up to 1 year [26]. Several putative mechanisms offer promise for the future [73, 74]. Energy intake can be modulated by decreasing energy absorption in the gastrointestinal tract either pharmacologically or surgically. Orlistat (Hoffman-LaRoche; Nutley, New Jersey) is a lipase inhibitor currently under evaluation in clinical trials. Energy stores can be reduced by adipose tissue lipolysis and by increased fatty acid oxidation. Finally, energy expenditure can be increased by exercise or thermogenic drugs [75]. Based on this broader understanding of the mechanisms underlying obesity, several new approaches to the treatment of obesity have been proposed [74] Table 5, of which the thermogenic drugs are among the most intriguing [76].

The ideal mechanism for the treatment of obesity could be one that increases energy expenditure and reduces food intake in the same manner as vigorous aerobic exercise. Thermogenic adrenergic agonists are among the most interesting candidates in this area because in some ways they mimic the effects of moderate amounts of exercise. Aerobic exercise is associated with various effects. First, it enhances overall cardiovascular fitness. It increases the potential for oxidation of fatty acids by muscle, and it slows bone loss by providing the stresses that strengthen trabecular bone. Exercise can also act as a feedback signal for modulation of food intake and for reduction of body fat stores. If one views norepinephrine and epinephrine as the major peripheral neurohormonal mediators for the effects of exercise, then thermogenic adrenergic agonists (which have these effects) might be desirable candidates for the treatment of obesity. Several thermogenic agonists are now under development. Ephedrine is the only currently available drug in this class. This drug can enhance energy expenditure in normal persons and, when used in combination with caffeine, can produce greater weight loss than placebo [75]. Of more interest, however, are the synthetic pharmacologic agents [76]. Only one of these agents has received a significant clinical trial [77]. In this study, the thermogenic agent produced greater weight losses than did placebo.

Limitations

Only two of the drugs that are currently available for treatment of obesity have had long-term trials. Because most of the older "noradrenergic" agents derived from amphetamine are no longer protected by patents, drug companies have little incentive to conduct long-term clinical trials with these agents. Thus, we have little information beyond the 12- to 20-week studies reviewed by Scoville [6]. With newer agents such as d,l-fenfluramine, D-fenfluramine, and fluoxetine, drug trials of 6 to 40 months have been published, and the drugs are clearly more effective than placebo. At present, only one trial that I know of has combined a serotonergic and noradrenergic drug [64]. As with antihypertensive medications, the use of combination therapy for obesity will probably produce a greater effect than any single agent alone. In addition to the lack of patent protection, several other barriers to the use of the current appetite-suppressing drugs exist [78, 79] (Table 6).

Public acceptance of new drugs for obesity contradicts the general perception that obesity is a disorder of willpower. If all a person needs to do is push themselves away from the table to keep from being fat, then why should drugs be used to treat this lack of willpower? This general public perception of the weak-willed obese person and the associated stigmatization that it carries are also found in the health profession. Health professionals and the regulatory agencies hold weight-reduction medications to a higher standard than that expected for other agents. Many conclude that a drug has failed if patients regain weight after termination of treatment. Because drugs have not "cured" obesity, one would not anticipate long-term maintenance of weight loss in most patients after cessation of treatment any more than one would expect blood pressure or cholesterol to remain low for long periods after termination of effective drug therapy. Thus, the attitudes of regulatory and health professionals need to change before new medications for obesity can be readily received. In addition, treatment for obesity—like treatment for hypertension or hypercholesterolemia—should be long-term. Clinical trials showing efficacy for 6 months and safety for 2 years should be sufficient to license a new drug to treat the obese diabetic and the obese hypertensive patient. Finally, the potential risks associated with weight cycling should not impair the use of therapy for obesity any more than the occasional stroke after cessation of hypertensive therapy should block the use of antihypertensive drugs. Where therapy is indicated, it should be undertaken with the long-term goal of maintaining weight reduction. Clearly, newer pharmacologic agents will be valuable adjuncts in this treatment goal.