Phentermine – Long-Term Use Study
Study Origin: The George Washington University Weight Management Program, Department of Medicine, George Washington University School of Medicine, Washington, DC. (Obesity Research 12:1821-1827 (2004))
Title: The Long-Term Management of Obesity with Continuing Pharmacotherapy
Researchers: Arthur Frank
Introduction: The search for useful pharmacotherapeutic agents (diet pills) for the management of obesity has been complicated by a series of therapeutic disasters. Dinitrophenol, thyroid, digitalis, amphetamines, and fenfluramine have successively cast a pall on the use of medications for weight management. Nevertheless, the recognition that eating and body weight are highly regulated biological phenomena, that obesity is a disease in which the regulation is dysfunctional, and that behavioral methods alone have limited effectiveness in enabling patients to achieve and sustain weight loss lead inescapably to the question of whether there should be some role for continuing use of medications in managing this disease.
It is very difficult to assess the usefulness of long-term pharmacotherapy. Individual anecdotal case reports are not useful for studying the problem; therapeutic recommendations cannot be based on the experience of one individual. No studies have documented the long-term safety of antiobesity medications for more than 2 years. The issue of whether any medication can or should be used as continuing, perhaps life-long, therapy cannot be resolved with standard clinical trials.
Because it may not be possible to do formal studies of very long-term use, there should be some value in examining the experience of individual patients who have utilized these medications in this form, particularly if a large number of patients with a similar clinical intervention can be assembled for a single assessment of their characteristics. The editors of the British Medical Journal have noted that “… case reports expose clinical issues that would otherwise escape attention”. A series of case reports can direct consideration to a pattern of similar events that have significant clinical importance. This analysis does not have the power of a clinical trial; it is merely a potentially useful clinical observation.
Eight patients are described here, each of whom has used weight loss medications for more than 10 years.
Clinical Material: Patient 1 (female, 52 years old ) has participated in our weight management program intermittently since 1977 (age 26). Her initial weight was 177 lb (80.5 kg, BMI = 32.4). With repeated effort, she was able to lose weight, but each time she was unable to sustain the weight loss. She reached a maximum weight of 192 lb (87.3 kg, BMI = 35.2). In 1993, she lost weight once again but gradually started to regain her lost weight. She then started phentermine to assist in the maintenance of her weight loss. She maintained her weight thereafter at <145 lb (65.9 kg, BMI <26.6). She regularly uses 15 to 30 mg of phentermine (usually 15 mg) >90% of the time. She occasionally misses a dose but notes more difficulty with eating when she does this. She briefly added fenfluramine and briefly substituted sibutramine, neither of which was helpful.
Patient 2 (male, 71 years old) weighed 215 lb (97.7 kg, BMI = 31.4) at age 17 in 1950. In about 1964, at age 31, he weighed about 205 to 210 lb (93.2 to 95.5 kg). He started medications for weight management and used either diethypropion or mazindol at different times thereafter. These were helpful in enabling him to lose weight and maintain his weight below 190 lb (86.4 kg, BMI < 27.7). He discontinued his medications when his physician died in 1988 and had more difficulty controlling his weight until he was seen in our program in 1992. He then started using phentermine when he weighed 193.25 lb (87.8 kg, BMI = 28.2). He continued using phentermine about 90% to 95% of the time and has maintained his weight below 185 lb (84.1 kg, BMI <27.0) since. He has added fenfluramine and substituted sibutramine, each without any additional value. He uses orlistat, with benefit, irregularly. He noted persistent, but tolerable, sleep and mood disturbances with diethylpropion and mazindol but no significant side effects with phentermine. In 2003 he had a resection of his prostate for prostatic hypertrophy and was found to have a small carcinoma of his prostate.
Patient 3 (female, 77 years old) lost weight deliberately at age 24 and used dextroamphetamine for weight control from the ages of about 28 to 38 years old. In 1981 at age 54, she was first seen in our program and weighed 177 lb (80.5 kg, BMI = 32.2). She lost weight, but over the next 10 years slowly regained about 75% of what she had lost and lost weight once again. In 1992, when she weighed approx. 155 lb (70.5 kg, BMI = 28.4), she started fenfluramine and phentermine but soon discontinued the fenfluramine. Initially, she used the phentermine inconsistently but then stabilized her pattern and rarely missed a dose thereafter. She lost approx. 20 lb (9.1 kg), regained a small amount, and stabilized her weight thereafter in the range of 140 to 150 lb (63.6 to 68.2 kg). She had mild and inconsistent blood pressure elevation that was controlled with hydrochlorothiazide and amlodipine. In 2003, at age 76, when she weighed 134 lb (60.9 kg, BMI = 24.4), she had a small cerebrovascular accident with full recovery and no sequelae. With great reluctance, she discontinued her use of phentermine. Six months later, her weight remained stable.
Patient 4 (male, 63 years old) weighed 196 lb (89.1 kg, BMI = 29.0) in 1986. He lost approx. 45 lb (20.5 kg) and was able to sustain most of that weight loss for approx. 3 to 4 years. He eventually regained his lost weight and reached a maximum weight of 205 lb (93.2 kg, BMI = 30.3). In 1993, he started phentermine and reached a weight of 195 lb (88.6 kg, BMI = 28.9) in the ensuing 10 years. He takes the medication continuously, rarely misses a dose, has never taken a holiday to test its effectiveness, and has never tried any other weight control medications. He had mild hypertension which resolved completely with this small weight loss.
Patient 5 (female, 55 years old) weighed 220 lb (100 kg, BMI = 35.0) in 1974. She lost approx. 80 lb (36.4 kg) but, with two pregnancies, controlled her weight irregularly during the next 5 years. In 1979, she was started on diethylpropion and, with few exceptions, has continued this medication for the ensuing 25 years. She has briefly tried other medications but has not found them to be helpful. She was first seen in our program in 1994 at which time she weighed 170 lb (77.3 kg, BMI = 27.1). She was continued on diethylpropion and lost approx. 35 lb (15.9 kg). Her weight has since fluctuated between 135 and 170 lb (61.4 to 77.3 kg). She uses 75 mg a day of diethylpropion and rarely misses a dose. Her most recent weight, in 2003, was 150 lb (68.2 kg, BMI = 23.7).
Patient 6 (male, 61 years old) weighed 222 lb (100.9 kg, BMI = 32.8) in 1969. After a modest weight loss, he started using phentermine to assist in the management of his weight. His weight since has fluctuated between 195 and 223 lb (88.6 to 101.4 kg). Although he has not been able to sustain any net weight loss, he is comfortable that the use of the medication has enabled him to avoid gaining additional weight. At times when he discontinued the medication, he had increased difficulty with careful eating and the maintenance of his current, albeit elevated, weight. He now rarely misses regular use of the phentermine. He briefly tried adding fenfluramine but found that it contributed little to his control with phentermine alone.
Patient 7 (male, 62 years old) was first seen as a patient for weight management in 1979. He then weighed 232 lb (105.4 kg; BMI = 30.7) but had previously reached a maximum weight of approx. 250 lb (113.6 kg, BMI = 33.1). During the following 15 years, he lost and regained significant amounts of weight eight times. In 1994, he successfully lost weight and started using phentermine for the final 10 kg of this weight loss effort. He continues to use this medication to assist in the maintenance of his lost weight. He uses it intermittently (approx. 50% to 60% of the time) because of uncomfortable symptoms of dysuria, urinary hesitancy, and nocturia. With this as-needed approach to the use of phentermine, he has been able (with two brief exceptions) to keep his weight below 215 lb (97.7 kg, BMI = 28.4). He has used methylphenidate for the management of attention deficit disorder. He does not use methylphenidate and phentermine simultaneously. His most recent weight, in 2003, was 181.5 lb (82.5 kg, BMI = 24.0).
Patient 8 (male, 36 years old) weighed over 300 lb (136.4 kg, BMI > 43.1) in 1990 and 266.25 lb (121 kg, BMI = 38.3) when he started our program in 1993. He lost 49 lb (22.3 kg). He then started phentermine to assist in his efforts to maintain his weight loss. He has used phentermine continuously during the ensuing 10 years. He also used fenfluramine approx. three to five times a week for approx. 3 to 4 years without significant additional benefit but discontinued this in approx. 1996. He used fluoxetine in 1995 and then bupropion in 1995 to 1996 for depression. His blood pressure has occasionally been slightly elevated but is usually normal. He takes the phentermine on most days. All of his recorded weights have been between 210 and 240 lb (95.5 to 109 kg, BMI = 30.2 to 34.5) in the past 10 years.
Results: Six of the eight patients used phentermine as the primary therapeutic agent. A seventh initially used diethypropion or mazindol and now uses phentermine. One has used diethylpropion regularly . The patients studied here occasionally used medications other than phentermine or diethylpropion, testing their effectiveness and tolerability, but all of the patients returned to the one medication that they have found to be most helpful. Many patients who use weight control medications do so intermittently or use them occasionally on an as-needed basis. Although this has been tested by most of these patients, intermittent use has not been the pattern, and all of the patients eventually settled into a pattern of using the medications in a relatively continuous way. Six of the patients used their medication >90% of the time of use.
There have been no complications associated with this long-term use and no significant adverse effects. This is reassuring considering the often-expressed concerns of many clinicians that side effects would be intolerable with continuing use. One patient limits his use because of dysuria. It should not be surprising, however, that these patients have not had significant side effects; those patients with medication-associated adverse events would not likely have continued to use the medications for this many years. No abnormal laboratory tests attributable to the medications have appeared. Blood pressure elevations have not occurred. Patients with pre-existing hypertension tended to normalize their blood pressure with weight loss. In none of these patients has there been an increase in blood pressure that could be temporally associated with the use of these medications. Anticipated adverse events (agitation, insomnia, nervousness, etc.) have not occurred and have not been more troublesome with the passage of time. The reverse appears to have been more common; start-up side effects tended to diminish, rather than increase, with time. None of the patients has had any evidence of dependency or addiction. There also has been no pattern of progressive decreased effectiveness and none of these patients has regularly, or even occasionally, increased the dose beyond ordinary therapeutic standards.
Seven of the eight patients were participants in a comprehensive weight management program with individual medical care, nutrition classes and counseling, behavior therapy, exercise therapy, psychoeducational groups, and psychotherapy. They used these support services variably, some quite extensively and some less so. All patients tended to decrease their use of these services as their weight stabilized. With long-term use, they were seen individually about every 3 to 12 months.
It is not possible to assess to what extent the usefulness of the medication can be associated with the comprehensive support services and if similar results could be obtained by an individual primary care physician working alone. In all cases, patients tended to use the support services more frequently when they were losing weight and then relied entirely on their relationship with the physician alone for the continuing use of the medication for the maintenance of their weight loss.
Some patients used the medication initially to help lose weight; some, initially, to maintain their weight. All ultimately used the medications to sustain their weight loss or to maintain a stable weight. One patient who had lost weight but did not sustain this weight loss was weight stable; he used the medication to avoid further weight gain. The effectiveness of this approach was verified for him by his reliable observation of weight gain when the medication’s effectiveness was tested by discontinuing its use.
Seven of the eight patients tested their use of their medication by episodes of discontinuing its use (taking a holiday), adjusting their dose, or trying alternate medications. All eventually abandoned these trials and settled with stable dose and schedule with a single drug.
Discussion: Many studies that have considered the long-term management of obesity have noted the relative ineffectiveness of current procedures and the need for continuing efforts to cope with the chronicity and intractability of its control. The issue of the long-term use of medications has been considered, but studies of this issue have been limited. Glazer has reviewed all long-term, placebo-controlled trials of obesity pharmacotherapy lasting >36 weeks. He focused particularly on the possibility of valvulopathy with this therapy. His review, although not sufficient to meet the standards of a meta-analysis, does, nevertheless, suggest the safety and efficacy of the available medications. Padwal et al.have conducted a meta-analysis of randomized, controlled, double-blind, weight loss and weight maintenance trials that lasted more than 1 year. The only medications for which adequate studies have been done that meet their inclusion criteria for a meta-analysis are orlistat and sibutramine. Both drugs appear modestly effective in promoting weight loss.
Many reports of short-term interventions of any sort are dismissed with suggestions that this kind of intervention offers little therapeutic benefit for the patient. The U.S. National Task Force on Prevention and Treatment of Obesity has noted that: “… obesity responds poorly to short-term interventions” and that “… there is little justification for the short-term use of anorexiant medications.” Participants in Weintraub’s landmark study of phentermine and fenfluramine “… had difficulty maintaining their weight loss without the anorexiant medications. Despite long periods of time at weights much lower than baseline, permanent resetting of the weight control mechanisms could not be demonstrated.”
The studies that have been done suggest that all of the currently available medications help some (but not all) patients lose a modest amount of weight (usually not >10% to 15% of their initial weight) during an initial 6 months of therapy. Continuing weight loss beyond that time is uncommon . This phenomenon of no further weight loss after about 6 months has been interpreted by some to be related to the ineffectiveness of the drug or the development of a tolerance to its therapeutic potency . Alternatively, this stabilization of weight loss may be thought to be similar to the effect of medications for other chronic diseases such as hypertension or hypercholesterolemia, the development of a therapeutic equilibrium. In this instance, the balance is maintained between the sustaining impact of the pharmacotherapy and the evolution of a series of counterregulatory mechanisms that defend body weight and protect the patient from starvation. This same phenomenon is similar in form, if not in magnitude, to the use of medications for hypertension, diabetes, and hypercholesterolemia.
Even a modest amount of weight loss is more readily sustained if the medications are continued than if they are stopped. Medications for obesity seem to be most effective if they are continued indefinitely, unless the weight is regained or significant side effects develop. In the absence of any better understanding of the mechanisms of the disease of obesity that will enable us to repair the dysfunctional settings of the eating regulation and body weight set-points, it seems possible that lifelong treatment may be necessary. The criterion for continuous use should be the clinical measure of its effectiveness, rather than a simple standard of duration of use.
What is obvious is that any pharmacotherapeutic intervention has no sustaining impact if the medication is discontinued. It appears that nothing about the mechanism of drug action permanently modifies the underlying metabolic abnormality. No drug cures the disease. The medication may continue to be helpful only if it is used in a continuous way. The short-term use of medications appears to be equivalent to the short-term use of medications for diabetes or hypertension. However, phentermine, the single most commonly prescribed drug for obesity treatment in the U.S. , is approved for “short-term” (12 weeks) use only. Its long-term use is not approved by Federal regulatory authorities in the United States.
Many other medical problems are managed with the continuing use of pharmacotherapy. There is no expectation that we will cure hypertension, hyperlipemia, or diabetes with medications. Rather, medications are used to establish control and to sustain that control. That concept, using medications to help manage rather than to resolve a medical problem, and using medications to sustain control, seems somehow to have been ignored in the management of obesity. There are a number of possible reasons for this therapeutic omission.
There is a cultural conviction that medications are inappropriate for obesity management. Many health professionals believe that eating is entirely a matter of choice and that losing weight depends simply on choosing to eat less. The history of the use of diet pills for obesity has been clouded by therapeutic complications, and many thoughtful physicians are fearful of being involved in another. Patients and medical personnel often have a view that the purpose of weight management is to lose weight. Once weight has been lost, by any method, there is a persistent belief that the issue has been resolved and little attention, and certainly no pharmacotherapy, is directed at the maintenance of weight loss. There is a fear that because medications have been approved by regulatory agencies for short-term use only, they would not be continuously effective. Because problems with abuse, dependency, and addiction occurred with previously used medications (amphetamines), there is a concern that these issues would arise with pharmacologically similar weight management drugs. There is concern that medications, which might be useful for morbid obesity, would be used carelessly, and at too great a risk, by patients with mild forms of this disease. Despite the recognition that some patients have derived continuing benefits from the long-term use of medications, others slowly regain their weight despite the medications. This has created the belief that weight gain is inescapable, regardless of any pharmacotherapy. Medical training, to the extent that it addresses the problem of obesity in any manner, tends to discourage the continuous use of any medication for weight management.
The primary medication used by the patients reported in this paper is phentermine, although two used diethylpropion, one initially and the other continuously. Other medications which have been available for many years are used less frequently in the U.S. Long-term use of amphetamine and its analogues is considered inappropriate because of concerns about dependency and abuse. Fenfluramine was little used until the publication by Weintraub et al. in 1992 of their studies of the long-term use of the fenfluramine-phentermine combination. It was withdrawn from the market (along with its analog, dexfenfluramine) in 1997 for safety reasons. Phenylpropanolamine and the caffeine/ephedrine combinations have, until recently, been available in the U.S. without a physician’s prescription. It is probable that any long-term use would not have involved physicians; therefore, its use could have passed without systematic note or study. Phendimetrazine and mazindol have not been used frequently in continuing management. The use of phendimetrazine has been limited by its classification as a category II controlled substance. Mazindol, no longer sold in the U.S., has had limited use, in some part because of its high cost. Sibutramine and orlistat, which seem to be effective and have been studied more extensively and for longer times than the others, have been available only for the past 5 years.
Long-term measures of effectiveness are difficult to assess. Control (nonmedication) patients were not studied here, but sufficient long-term experience has established that most patients are not ordinarily able to sustain their weight loss. Clearly, the patients studied here have self-selected this particular therapeutic format and have been successful with its use. Patients who do not respond well, or who have adverse effects, would not be expected to continue its use. It is not possible, by studying the experience of these patients, to establish any generalizations about the usefulness of this approach. Rather, it is important to demonstrate that some patients will find this form of therapy useful and that some consideration should be given to the identification of these patients and the use of long-term pharmacotherapy for them.
Some of the current medications for the treatment of obesity have been available for >30 years, but systematic long-term studies of their utility have not been published. This is, in fact, true for most medications. There is no realistic way that long-duration (more than 2 years) studies for this category of medication can be blinded and placebo-controlled. Efficacy of long-term use of any medication depends on other assessments; collection (usually by pharmaceutical manufacturers or government agencies) of adverse events and the clinical assessment of whether or not the drug in question sustains its therapeutic usefulness. It is obviously true that we will require 10 years to establish the safety and efficacy of 10 years of treatment with any pharmacotherapy for any medical problem.
Although this generalizable pharmacotherapeutic dilemma seems obvious, medications for the treatment of obesity seem to have been held to a higher standard of proof of efficacy than medications for other therapeutic categories. This seems, in part, to be related to many factors:
the complications observed with previously used obesity medications; the relatively small therapeutic effect that these medications have (usually the loss of <10% to 15% of the patient’s initial weight); the patient’s and the physician’s belief that 10% to 15% weight loss is therapeutically insignificant for patients with substantial obesity and the belief that this magnitude of weight loss is insufficient to justify the use of medications; and the long-standing cultural perception that obesity is a trivial problem, that it is caused by the patient’s willful misconduct, that it should be easily treated by the patient’s behavior change, and that, therefore, no pharmacotherapeutic risk is tolerable.
Given these beliefs, the long-term use of medications for weight management has often been dismissed as therapeutically inappropriate. Moreover, there is an additional cultural bias which directs most therapeutic efforts to weight loss, rather than to the maintenance of weight loss. From this naturally derives an astonishingly naive view that once the patient has completed weight loss, any medication can (or should) be discontinued. This therapeutic misunderstanding would be intolerable for the control and sustaining management of comparable chronic medical problems such as diabetes, hypertension, or hypercholesterolemia.
Individual physicians who have prescribed medications in a continuous form note that their patients use them intermittently (perhaps systematically, as on weekends or when traveling) or on an as-needed basis, when they start to regain after a period of some weight stability. As-needed use of medications, particularly if by patient choice, is a common therapeutic option for many medical problems; arthritis, asthma, anxiety, and chronic pain are obvious examples. It is very likely that there is a useful role for this approach for long-term weight management as well. Most of the patients studied here have tried this kind of medication use, but all have settled down to a relatively stable and continuous pattern of use. Missed doses are usually inadvertent, not deliberate.
Similarly, many of these patients have tried alternate medications and/or have tried a combination of medications. This combination use is limited because there are relatively few current therapeutic options, the use of many therapeutic combinations may be pharmacologically contraindicated, and the cost of multiple medications (particularly in the U.S. without insurance reimbursement) is substantial. One of these patients currently uses orlistat advantageously with phentermine. The pattern of switching among therapeutic options or combination therapy has not evolved with these patients.
Many evaluations of the use of weight management medications have noted the very low incidence of addiction, dependence, abuse. or inappropriate use. Although caution, vigilance, and patient monitoring are clearly necessary for continuous use of these medications, the nonabusive patterns in these patients and the noteworthy absence of side effects are therapeutically reassuring.
Although in all cases the medications were prescribed to help patients lose weight or maintain weight loss, one patient has been unable to sustain the weight loss and has used the medication successfully to maintain a controlled and stable, albeit moderately elevated, weight. This invariably raises the question of prevention of weight gain and the substantial difficulty in establishing if any medication is truly helpful for this purpose. It may be possible to justify this prevention use in a patient with a BMI >30 but it raises the awkward problem of how medical professionals can cope with the problem of mildly overweight patients. All physicians have encountered the patient who is only slightly overweight, who has no evidence of an eating disorder, but who struggles to avoid gaining weight because of a bad family history and the evolving characteristics of the metabolic syndrome. Both the physician and the patient recognize the ominous prognosis; all patients who are substantially obese were, at one time, merely a little overweight. Surely the issue of continuous medication will be raised in the consideration of the management of these patients, but nothing in the analysis of the experience of these long-term patients addresses the appropriateness of prevention, particularly for patients who are mildly overweight.
It is noteworthy also that many of these patients lost weight without medications (diet pills) and did not use the medications primarily to lose weight. Perhaps a neglected role for the medications can be for the greater and more intractable task of weight maintenance, rather than for the often more accessible problem of weight loss. It may be that the possibility and opportunity for medications for maintenance will be greater than their role in weight loss. Weintraub et al., in their papers on the use of phentermine and fenfluramine, addressed the concept of long-term use of weight management medications and the effectiveness of medications for this purpose. Their studies emphasized the consideration of the continuous management of obesity as a chronic disease. It is regrettable that this concept has not been systematically evaluated with additional studies.
It is likely that more medications will be available in the next decade. Some consideration must be given to the important question of their potential efficacy for long-term use.