Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
International Journal of Obesity (2007) 31, 850–857 & 2007 Nature Publishing Group All rights reserved 0307-0565/07 $30.00 www.nature.com/ijo ORIGINAL ARTICLE Comparison of combinations of drugs for treatment of obesity: body weight and echocardiographic status LD Whigham, NV Dhurandhar, PS Rahko and RL Atkinson Departments of Medicine and Nutritional Sciences, University of Wisconsin, Madison, WI, USA Background: Obesity treatment with single drugs produces weight losses of about 8–10% of initial body weight. Few studies of combinations of drugs for treating obesity have been published. The combination of phentermine, an adrenergic agent, and fenfluramine, a serotonergic agent, (phen–fen) produced weight losses of about 15% of initial body weight. Fenfluramine is no longer available because it was associated with cardiac valve lesions. Phentermine–fluoxetine (phen–flu) has been proposed as an alternative for phen–fen. Objective: To compare the efficacy of treatment and prevalence of cardiac valve abnormalities on phen–flu vs phen–fen. Design: Retrospective chart review of all patients treated for at least 3 months with phen–flu (N ¼ 97) to a random sample of patients treated with phen–fen (N ¼ 98) in the Clinical Nutrition Clinic at the University of Wisconsin. Comparison of echocardiograms in all patients treated solely with phen–flu (N ¼ 21) to a random sample of patients treated with phen–fen (N ¼ 47), and to a group of subjects never treated with obesity drugs (N ¼ 26). Results: With last observation carried forward analysis (LOCF), at 6 months of treatment the phen–fen patients lost 12.670.6% of baseline weight and phen–flu patients lost 9.070.6% (Po0.001). With completers analysis, there were no significant differences in weight loss as a percent of baseline weight at 6 months (14.470.6 vs 13.370.9%). LOCF decreases in body mass index (BMI) at 6 months were 5.3 and 3.6 kg/m2 for phen–fen and phen–flu, respectively (Po0.001), and 6.270.3 vs 5.470.4 kg/m2, respectively, for the completers analysis (P – NS). Dropout rate at 6 months was higher in phen–flu subjects (44 vs 28%). In subjects without atherosclerosis of valves (presumably pre-existing), cardiac valve lesions occurred in eight of 38 phen–fen subjects and in none of 15 phen–flu subjects or 25 control subjects who had not been treated with drugs. Conclusions: The combination of phentermine and fluoxetine was not as effective as phen–fen, but was not associated with cardiac valve lesions. Longer term, larger scale studies of phen–flu are warranted. International Journal of Obesity (2007) 31, 850–857. doi:10.1038/sj.ijo.0803498; published online 5 December 2006 Keywords: fluoxetine; fenfluramine; phentermine; cardiac valves; weight loss; co-morbidities Introduction The use of drugs for the treatment of obesity was discouraged by most physicians for several decades from the 1960s into the early 1990s.1 Obesity drugs were thought to be ineffective and dangerous. The US Food and Drug Administration (FDA) approved the use of obesity drugs for up to about 12 weeks, and many states had laws prohibiting the use of obesity drugs at all, or severely curtailing their use.1 The minority of physicians who prescribed obesity drugs usually followed the recommendations for short-term use. Only single drugs were prescribed and use of combinations was virtually unknown.1,2 The seminal series of reports by Correspondence: Dr RL Atkinson, Obetech Obesity Research Center, 800 E. Leigh St., Suite 50, Richmond, VA 23219, USA. E-mail: [email protected] Received 30 August 2005; revised 17 July 2006; accepted 13 September 2006; published online 5 December 2006 Weintraub et al.2 in 1992, demonstrating that the combination of phentermine and fenfluramine (phen–fen) could produce clinically significant weight loss for periods of up to 3.5 years, ignited the ‘phen–fen revolution’ that saw thousands of physicians prescribing this combination to millions of patients. Subsequent studies confirmed longterm weight losses that were superior to diet and exercise alone and to treatment with single drugs.3–9 Weight losses at 1 year were approximately 15–17% of initial body weight with phen–fen3–9 compared to losses of 6–10% of initial body weight with single-drug treatment using fenfluramine.10–19 When fenfluramine and dexfenfluramine were removed from the market in September 1997 because of concerns about cardiac valve lesions,4,6,20–33 some physicians turned to combinations of phentermine and other serotonin-active drugs such as fluoxetine and sertraline in an attempt to reproduce the phen–fen effect. However, there are very few Combinations of obesity drugs LD Whigham et al 851 studies published on the efficacy of these combinations for obesity, and none have compared results with phen–fen. Also, there are minimal data about the safety of selective serotonin reuptake inhibitors (SSRI) when used in combination with phentermine. Rowland et al.34 reported in animals that the combination of phentermine and fluoxetine (phen–flu) had additive effects on weight loss. Anchors35 published a book for the lay press proposing the use of phen–flu, but did not perform controlled studies. Devlin et al.36 treated overweight binge eaters with phen–flu for 20 weeks and found a reduction of 9% of initial body weight and a reduction in binge eating. These effects did not persist when the medications were stopped. Reeve et al.37 demonstrated that phentermine and fluoxetine did not appear to increase the risk of pulmonary hypertension. Bostwick and Brown38 reported a single case of toxicity with the use of phen–flu. In an uncontrolled case series, Griffen and Anchors39 reported in a letter that phen–flu was not associated with cardiac lesions on echocardiography. The objectives of this study were to evaluate the efficacy and safety of treatment with the combination of phen–flu, and to compare to treatment with the combination of phen–fen. We evaluated changes in body weight, blood pressure and selected laboratory tests, and the occurrence of cardiac valve lesions on echocardiography with the two combinations. with cardiac valve lesions, we were concerned that the combination of phen–flu might have a similar association. We contacted all the subjects who had taken or were taking phen–flu and invited them to undergo an echocardiogram. Of the subjects who agreed to have an echocardiogram, 21 had never taken fenfluramine or dexfenfluramine, and had been treated with phen–flu for at least 3 months. The average time of treatment with phen–flu was 5137310 days (7s.d.). As a comparison group, randomly selected subjects from the University of Wisconsin CNC population who had been treated with phen–fen for at least 3 months and had not ever taken fluoxetine were invited to obtain an echocardiogram. A total of 47 subjects underwent an echocardiogram, so all were included. The average time of treatment with phen–fen was 4457216 days (7s.d.). Finally, an additional comparison group of 26 subjects from the University of Wisconsin CNC research population who had never taken fenfluramine or any selective serotonin reuptake inhibitor medication was invited to have an echocardiogram and acted as a negative control group. Treatment protocol, efficacy evaluation An initial history was taken and physical examination was carried out on all subjects and an electrocardiogram and Table 1 Materials and methods Subjects, efficacy evaluation In a retrospective chart review, all patients who had received phentermine and fluoxetine in the Obesity Treatment Program of the Clinical Nutrition Clinic (CNC) of the University of Wisconsin between 1994 and 1998 were identified. A second inclusion criterion was that none of these subjects had ever taken fenfluramine or dexfenfluramine before the study period. In some instances, subjects who took phen–flu had been precluded from taking phen– fen because of a history of depression, which was an exclusion criterion for phen–fen in our clinic. A total of 97 phen–flu subjects were identified. A comparison group of 98 patients was selected randomly from over 700 patients in the CNC program who had undergone treatment with the combination of phen–fen, but who had never taken fluoxetine or other selective serotonin reuptake inhibitors. Because of the exclusion criteria, none of these subjects had a history of depression. Table 1 describes the inclusion and exclusion criteria and Table 2 describes the baseline characteristics of the subjects in the two groups. Subjects, echocardiogram evaluation When fenfluramine and dexfenfluramine were taken off the market in September, 1997, because of an association Inclusion and exclusion criteria for drug treatment Inclusion criteria 1 Body mass index (kg/m2) greater than 27. 2 Committed to area for at least 2 years. Exclusion criteria 1 Age greater than 65 years or less than 18 years. 2 Pregnancy or lactation 3 Malignant arrhythmia, CHF, or unstable cardiac disease. 4 Uncontrolled hypertension – diastolic 4110, systolic 4195 mm Hg. 5 Presence of any severe systemic illness. 6 At risk of increased protein catabolism (disease or drugs). 7 Other drug therapy if there is interaction with study drugs. 8 History of psychosis, bipolar disorder, or anorexia nervosa. 9 Presence of major active depression. Abbreviation: CHF, chronic heart failure. Table 2 Baseline characteristics of subjects in the efficacy evaluation Phen–fen s.e.m. Phen–flu s.e.m. P-value Age (year) Weight (kg) BMI (kg/m2) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Glucose (mg/dl) Total cholesterol (mg/dl) Triglycerides (mg/dl) HDL (mg/dl) 44.0 120.7 42.3 133.5 85.4 103.6 213.7 167.1 43.8 71.04 73.6 71.03 71.91 71.16 73.30 73.75 711.63 71.22 44.8 70.93 0.6 107.1 72.6 0.002 38.7 70.80 0.005 128.6 71.78 0.06 82.6 70.89 0.056 100.1 72.73 0.4 210.9 73.95 0.6 153.4 710.76 0.4 47.9 71.30 0.02 Abbreviations: BMI, body mass index; HDL, high-density lipoprotein. International Journal of Obesity Combinations of obesity drugs LD Whigham et al 852 standard laboratory tests were performed. The laboratory tests included a complete blood count, chemistry-20 panel, lipid panel, magnesium, thyroid function tests, urinalysis and pregnancy test in fertile females. Then, all subjects were prescribed a diet designed to provide about 500–1000 kcal/ day lesser than their usual intakes, calculated from 7-day food diaries. The calorie deficit was determined by initial size, personal preference and desired rate of weight loss. An exercise/activity program consisting of walking or comparable activity for 30 min or more, three or more times weekly, was prescribed. At each visit, subjects filled out a follow-up questionnaire listing symptoms, complaints and compliance to the diet and exercise program. Subjects who did not adhere to the diet and/or exercise program were encouraged to do so at each visit, but no penalties were incurred for failure to adhere. Subjects were prescribed medications starting with low doses and increasing if weight loss was unsatisfactory. Phentermine was started at half of a 37.5-mg tablet, or 18.75 mg per day. Fenfluramine and fluoxetine were started at 20 mg per day. A weight loss of less than one pound per week, averaged over 4 weeks was used as the criteria to go to the next highest dose. The usual pattern of stepwise increase was to raise fenfluramine to 40 mg per day, then to 60 mg per day, followed by increasing phentermine to 37.5 mg per day. On rare occasions, phentermine would be increased to 1.5 tablets per day (56.25 mg) if the patient had responded to the initial increase in phentermine. Except in very rare instances, the maximum doses of fenfluramine and fluoxetine were 60 mg per day. Patients returned to the CNC every 2 weeks for 24 weeks, then were scheduled to be seen every 4 weeks. At each visit, patients were seen by the dietitian and the physician or nurse practitioner for individual evaluation, examination of follow-up questionnaire and food diaries, and refilling of prescriptions when necessary. For most of the sessions, a group lesson was also given. Group lessons lasted about 30 min and dealt with some aspects of obesity or its treatment, as previously reported by our group.40 Research design, efficacy evaluation Measures of efficacy included percent change from initial body weight and changes in body mass index (BMI), blood pressure, serum glucose, serum cholesterol, triglycerides and high-density lipoprotein (HDL). Last observation carried forward (LOCF) analyses were performed on the variables of change in BMI and percent change in body weight. Also, analyses were performed on subjects who completed 6 months of treatment. Research design, echocardiogram evaluation Echocardiograms were performed in the Cardiology Laboratory of the University of Wisconsin Hospital using an Agilent Technologies, model 5500 Andover (MA, USA) apparatus. International Journal of Obesity The co-author, Dr Peter Rahko, read the echocardiograms blinded to the subject group. Subjects who had calcifications in the aortic valve or in the aortic root were discarded from analysis from all groups because atherosclerosis is not a feature of phen–fen-associated valvulopathy, but may produce valvular insufficiency. Valvulopathy was rated in two ways. First, numerical scores were assigned to the severity of valvular insufficiency: 0 – none, 1 – trace, 2 – mild, 3 – moderate, 4 – severe. Second, ‘significant’ valvulopathy as defined by the US FDA was identified.21,41 The FDA designated mild or greater aortic insufficiency, and moderate or greater mitral insufficiency, as serious valvular insufficiency. The number of subjects in each group who had serious valvulopathy by these criteria was recorded. Statistical analyses The obesity treatment data were analyzed using both an LOCF analysis and a completers analysis (completing 6 months of treatment). The two drug groups were compared by the non-paired Student’s t-test. The echocardiographic data were compared by analysis of variance (ANOVA), followed by Fisher’s least significant difference (LSD) (protected t-tests) for the numerical scoring of valvular disease, and by non-parametric analysis, Kruskal–Wallis one-way ANOVA, for ‘significant’ valve disease. Software used for the statistical analyses were Excel (Microsoft, 1997) and GB Stat, Version 6.5 (Dynamic Microsystems, Silver Springs, MD, USA). The protocols and informed consent forms were approved by the Human Subjects Committee of the University of Wisconsin, Madison, Medical School. Results Changes in body weight and BMI The data were analyzed using both an LOCF analysis and a completers analysis (completing 6 months of treatment) (Figure 1). Regardless of the method of analysis, differences from baseline to 6 months were highly significantly different in both groups (Po0.01). In the LOCF analyses, body weight decreased in the phen–fen group from 120.573.6 to 105.47 3.2 kg and BMI decreased from 42.371.0 to 37.070.9 kg/m2. In the phen–flu group, body weight decreased from 107.27 2.6 to 97.372.3 kg and BMI decreased from 38.870.8 to 35.270.7 kg/m2. The phen–fen group weight loss of 15.170.9 kg was significantly greater than the weight loss of 9.970.8 kg in the phen–flu group (Po0.0001). When expressed as percent change from baseline, phen–fen lost 12.6% and phen–flu lost 9.0% (Table 3). A completers analysis was performed on the 71 phen–fen subjects (72%) and the 44 phen–flu subjects (45%) who completed 6 months of treatment. Body weight decreased in the phen–fen group from 126.074.3 to 108.373.9 kg and BMI decreased from 44.171.3 to 37.971.2 kg/m2. In the Combinations of obesity drugs LD Whigham et al 853 phen–flu group, body weight decreased from 112.074.4 to 97.274.1 kg and BMI decreased from 40.071.3 to 34.77 1.2 kg/m2. When expressed as percent change from baseline, phen–fen lost 14.4% and phen–flu lost 13.3% (Table 3). None of the differences between groups were significant for the completers analyses. Evaluation of individuals who remained in treatment for 1 year showed that phen–fen patients lost 18.571.2% of initial body weight and phen–flu patients lost 15.471.8% of initial weight. The difference in weight loss between groups was not significant, but dropout rate was 45% with phen–fen and 82% with phen–flu. a 130 Kilograms (kg) 125 120 Responses of blood pressure and laboratory assays Systolic blood pressure decreased significantly from baseline to 6 months in both phen–fen and phen–flu groups (Tables 4 and 5), but there was no significant difference between groups. Diastolic blood pressure dropped significantly in the phen–fen, but not in the phen–flu group, and the difference between groups was not significant (Tables 4 and 5). Serum glucose, cholesterol and triglycerides all decreased significantly from baseline to 6 months in both groups (Tables 4 and 5). Serum cholesterol decreased by 32.973.1 mg/dl in the phen–fen group, vs a decrease of only 9.274.1 mg/dl in the phen–flu group (Po0.0001). However, HDL decreased by 3.071.12 mg/dl (Po0.001) in the phen–fen group compared to a non significant increase of 1.171.4 mg/dl for the phen– flu group. The difference in change in HDL was significant between groups (Po0.01). 115 Subject dropouts The average time in treatment for phen–fen subjects was 393 days, and was 226 days for phen–flu subjects (Po0.0001). 110 105 100 95 Table 4 90 Baseline Month 3 b 50 Phen–fen patients Glucose (mg/dl) Total Cholesterol (mg/dl) Triglycerides (mg/dl) HDL (mg/dl) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) 45 BMI (kg/m2) Laboratory and blood pressure values from baseline to month 6 Month 6 40 35 30 Baseline Month 3 Month 6 Figure 1 Changes in body weight (a) and BMI (b) during 6 months of treatment, last observation carried forward. Circle ¼ phen–fen; trianglephen–flu. Table 3 Change in body weight and BMI for last visit carried forward (LOCF) analysis and completers analysis Phen–Fen s.e.m. LOCF Change in Weight (kg) 15.1 70.9 Percent Change in Weight 12.6% 70.6% Change in BMI 5.3 70.3 Phen–Flu s.e.m. s.e.m. 105.27 213.27 73.63 95.11 74.36 180.36 72.32 o0.0001 73.67 o0.000001 154.07 45.56 136.48 79.35 101.01 71.56 42.49 72.33 129.34 75.64 o0.000001 71.54 o0.001 72.35 o0.0001 71.45 71.24 o0.001 86.56 101.15 205.20 Month 6 82.34 73.62 95.05 76.90 196.03 s.e.m. P-value 72.67 o0.05 77.10 o0.05 151.12 715.41 110.10 710.84 o0.0001 47.02 72.10 48.15 72.07 NS 131.21 72.77 124.70 72.02 o0.05 82.64 71.31 81.30 71.40 NS Abbreviations: HDL, high-density lipoprotein; NS, not statistically significant. P-value 9.9 70.8 o0.001 9.0% 70.6% o0.001 3.6 70.3 o0.001 Completers Change in Weight (kg) 17.7 71.0 14.8 71.2 Percent Change in Weight 14.4% 70.6% 13.3% 70.9% Change in BMI 6.2 70.3 5.4 70.4 Phen–flu patients Glucose (mg/dl) Total Cholesterol (mg/dl) Triglycerides (mg/dl) HDL (mg/dl) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Baseline NS NS NS Abbreviations: BMI, body mass index; LOCF, last visit carried forward; NS, not statistically significant. Table 5 Changes in laboratory and blood pressure values after 6 months of treatment Phen–fen s.e.m. Phen–flu s.e.m. Glucose (mg/dl) Total Cholesterol (mg/dl) Triglycerides (mg/dl) HDL (mg/dl) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) 10.2 32.9 53.1 3.0 7.1 4.2 P-value 72.5 6.1 72.7 NS 73.1 9.2 74.1 o0.0001 77.7 41.0 79.8 NS 70.9 1.1 71.4 o0.01 71.8 6.5 72.8 NS 71.3 1.4 71.3 NS Abbreviations: HDL, high-density lipoprotein; NS, not statistically significant. International Journal of Obesity Combinations of obesity drugs LD Whigham et al 854 Dropout rate in the phen–fen group was 9% at 3 months, 28% at 6 months and 44% at 1 year, whereas it was 29, 56, and 82%, respectively in the phen–flu group. The most common reasons for dropouts in the phen–fen group were discontinuation of program (42%) and removal of the drug from the market (27%). The most common reason for dropout from the phen–flu group was a desire to switch medications (44%), virtually always to phen–fen. Dropouts due to side effects were 13% in each group. Eight percent of subjects dropped out of the phen–flu group for slow or insufficient weight loss. Only 1% dropped out of the phen–fen group for insufficient weight loss. Echocardiogram evaluation Subjects in each of the three groups who had calcifications in the aortic or mitral valves or in the aortic root were discarded from the analysis (phen–flu: six of 21; phen–fen: nine of 47 and control: one of 26). Remaining subjects were compared by ANOVA using scores for severity of valvular disease and by nonparametric analysis for presence or absence of FDA classification of valvular disease, as described above. Fisher’s LSD (protected t-tests) after ANOVA revealed that subjects on phen–fen had significantly higher severity of aortic insufficiency valvular disease (0.4770.14) than did phen–flu (0.0070.00) or control (0.1270.07) (Po0.05). There were no significant differences in severity score for mitral insufficiency valvular disease among the three groups (Table 6). Using FDA criteria for significant valve disease, eight of the 38 subjects (21%) on phen–fen had ‘significant’ aortic disease and one had significant mitral disease. None of the phen–flu or control subjects had significant valve disease. However, Kruskal–Wallis one-way ANOVA revealed no significant differences among the three groups for either aortic or mitral valve disease. Discussion Obesity is a chronic disease that requires continuous, probably lifelong treatment.42 Like other chronic diseases, if treatment stops, the results of treatment are lost, and weight is regained in the vast majority of patients. Virtually all other chronic diseases are treated with pharmacologic agents, and it seems likely that the standard treatment of obesity eventually will include drugs. Unfortunately, few chronic diseases can be treated with single drugs, and combinations of drugs are required for most. However, current FDA recommendations and manufacturers’ recommendations generally discourage the use of combinations of obesity drugs. The major contribution of Weintraub’s studies with phentermine and fenfluramine were to demonstrate that a combination of two obesity drugs was more effective than either drug alone, and that long-term treatment produced continued weight loss.2 International Journal of Obesity Table 6 Severity scores for valvular disease on echocardiograms (7s.e.m.) Groups Control Phen–fen Phen–flu Aortic insufficiency Mitral insufficiency 0.1270.07 0.4770.14a 0.0070.00 0.8070.16 0.8470.15 0.6770.16 a Po0.05 compared to control and phen–flu groups. The unexpected finding of cardiac valve lesions associated with fenfluramine and dexfenfluramine resulted in the removal of these drugs from the market.4,6,20–33 This left the field of obesity with only three classes of FDA-approved obesity drugs: adrenergic agents (e.g. phentermine, diethylpropion, phendimetrazine), a combined adrenergic–serotonergic agent (sibutramine), and a malabsorptive agent (orlistat). There are very limited data on the subject, but combinations of either sibutramine and orlistat43 or phentermine and orlistat44 do not appear to increase weight loss over the use of single drugs. Thus, there are no accepted or commonly used combinations of obesity drugs at this time. Fluoxetine has been studied extensively as a potential weight reduction agent, and has been shown to cause significant weight loss after 6 months of treatment.45,46 However, unlike other drugs approved for the treatment of obesity, weight losses did not persist, and after 1 year of treatment body weight was back to baseline.45,46 Longer term studies suggest that fluoxetine, when used alone, may cause significant weight gain over time.47 Data on phen–flu treatment are minimal. An uncontrolled case series reported in the lay press that the combination of phen–flu had a greater effect on weight loss than did the use of single drugs alone.35 In an open, uncontrolled case series Devlin et al.36 noted that phen–flu reduced body weight and incidence of binge eating at 20 weeks, but these benefits were lost when the medications were stopped. As both obesity and binge eating disorder are chronic diseases, it is not surprising that the response stopped when treatment was stopped. In virtually all other chronic diseases, when effective treatment is stopped, the disease returns. This has been amply demonstrated for obesity as well,42 so the conclusions of these authors that, ‘this study does not support the long-term clinical utility of adding phentermine/fluoxetine to cognitive-behavioral therapy for binge eating disorder’, does not appear justified. The current study extends the preliminary, uncontrolled observations of Anchors35 and Devlin et al.36 and demonstrated that the combination of phen–flu produced a weight loss of 13.3% of initial weight at 6 months of treatment in subjects remaining on treatment. This loss was not different from the 14.4% weight loss achieved with phen–fen in completers at 6 months. The LOCF analyses present a different picture of the effectiveness of phen–flu. Weight loss at 6 months was 9.0% compared to 12.6% for phen–fen, a highly significant difference (Po0.001). It is likely that phen–fen is more effective than phen–flu owing to the Combinations of obesity drugs LD Whigham et al 855 nature of the drugs, but subjects on phen–flu also had other variables that may have affected the difference in weight loss. The phen–flu subjects were 13.3 kg lighter. Weight loss tends to be greater in heavier individuals. Also, some of the phen–flu subjects had a history of depression or had been treated with antidepressants in the past, and this precluded treatment with phen–fen in our clinic. Therefore, the populations may have been somewhat different at the baseline. The dropout rate was greater in the phen–flu subjects. For example, at 1 year, only 18% of the subjects remained on phen–flu, compared to 56% on phen–fen. The differential in the dropout rate may be partially explained by the fact that these studies were conducted at the height of the interest in phen–fen, and many of our patients asked to be switched from phen–flu to phen–fen, despite significant weight loss. As subjects who had lost less weight were more likely to drop out, this makes the phen–flu weight loss for completers look better. A limitation of this study is that neither of the combinations was compared to single-drug treatment. There have been numerous studies conducted with all of the obesity drugs on the market, and the weight losses average about 5–10% of initial body weight at 6 months to 1 year. An exception was the Sibutramine Trial of Obesity Reduction and Maintenance (STORM) trial with sibutramine,48 carried on in multiple sites in Europe, which achieved a loss of about 13% of initial body weight, and a study by Wadden et al.49 that demonstrated a weight loss of 16.5% at 1 year with sibutramine, rigid diet and lifestyle modification. However, the study by Wadden et al.49 and most of the sites in the STORM trial used very low-calorie diets to achieve a large initial weight loss, so the data are not comparable to those of other trials. Our data, and the data of Anchors,35 suggest that the weight losses achieved with phen–flu may be higher than those usually achieved with single obesity drugs currently on the market, and that this combination therapy may be justified. As would be expected, both phen–flu and phen–fen reduced the complications of obesity, including blood pressure, serum cholesterol and triglycerides. Phen–fen had a much greater effect on lowering serum cholesterol than did phen–flu (Po0.001). In contrast, phen–flu produced a slight increase in serum HDL that was significantly different (Po0.01) from the reduction in HDL seen with phen–fen. Thus, the improvements in cardiovascular risk with weight loss from both of the combinations appear approximately comparable. These improvements in cardiovascular risk profile may justify treatment with obesity drugs, even if total weight loss falls far short of what patients desire.50 The numbers of subjects who underwent echocardiography are relatively small, and our data may be subject to a type 2 error. Nevertheless, there was a difference between the groups in one of the methods used to assess cardiac valve lesions. Using a scoring system for the degree of valve damage, phen–fen had a significantly greater aortic valve damage score than did either of the other groups. However, when using the FDA scoring system of ‘significant’ aortic or mitral valve lesions, there were no significant differences, probably owing to inadequate power. Neither phen–flu nor control subjects had any FDA ‘significant’ lesions, whereas eight of 38 phen–fen subjects had such lesions. Despite the extensive use of fluoxetine for long-term treatment of depression, there have been no reports of cardiac valve lesions similar to those seen with fenfluramine and dexfenfluramine.4,6,20–33 Rothman et al.51 suggested that serotonergic drugs that bind to 5-HT (2B) receptors were associated with cardiac valve lesions. Fenfluramine and dexfenfluramine were among those that bound to 5-HT (2B) receptors, but fluoxetine did not.51 Reeve et al.37 evaluated the effects of phentermine and fluoxetine on pulmonary arterial pressure and electrophysiology, as the fenfluramines had been demonstrated to be associated with primary pulmonary hypertension.52,53 In contrast to dexfenfluramine, fluoxetine and phentermine had minimal effects on the cell resting membrane potential, and Reeve et al.37 concluded that they were unlikely to be associated with pulmonary hypertension. During the initial stages of the alert regarding cardiac valve lesions associated with obesity drugs, several authors speculated that the addition of phentermine might have exacerbated the association of fenfluramine and dexfenfluramine with valve lesions.20,21,25,54 However, there is no credible evidence that this is so, and the FDA has not limited the use of phentermine. Phentermine was not associated with primary pulmonary hypertension, in contrast with the fenfluramines.53 Our data do not provide assurance that phen–flu will not be associated with cardiac valve lesions, but they do suggest that any effects would be much smaller than those associated with fenfluramine, dexfenfluramine, or phen–fen. However, any practitioner who chooses to use the combination of phen–flu would be wise to provide extensive information and obtain informed consent, preferably written. In summary, the combination of phen–flu produced weight losses that were not as great as those achieved with phen–fen when LOCF analysis was used. There were no differences in weight loss between the two combinations in patients who completed 6 months of treatment, but there was a much higher dropout rate in the phen–flu group. Blood pressure and serum lipids demonstrated significant improvements with both combinations. The weight losses and improvements in cardiovascular risk achieved in our trial give encouragement that additional, larger trials with phen–flu combination would be useful. Future trials should compare the combination of phen–flu with the results of phentermine and other obesity drugs used alone. Acknowledgements We thank Dr E Shrago, H Walaski, E Freitick, and A Rettammel for clinical assistance; and J Hansen and V International Journal of Obesity Combinations of obesity drugs LD Whigham et al 856 Schmidt for administrative assistance. Supported by grants from the Eli Lilly Company and funds from the Beers– Murphy Clinical Nutrition Center, University of Wisconsin, Madison. References 1 Bray GA. Drug treatment of obesity. Am J Clin Nutr 1992; 55 (2 Suppl): 538S–544S. 2 Weintraub M. Long-term weight control: the National Heart, Lung, and Blood Institute funded multimodal intervention study. Clin Pharmacol Ther 1992; 51: 581–646. 3 Atkinson RL, Blank RC, Schumacher D, Dhurandhar NV, Ritch DL. Long-term drug treatment of obesity in a private practice setting. Obes Res 1997; 5 : 578–586. 4 Wadden TA, Berkowitz RI, Silvestry F, Vogt RA, St John Sutton MG, Stunkard AJ et al. The fen–phen finale: a study of weight loss and valvular heart disease. Obes Res 1998; 6: 278–284. 5 Dhurandhar NV, Blank RC, Schumacher D, Atkinson RL. Initial weight loss as a predictor of response to obesity drugs. Int J Obes Relat Metab Disord 1999; 23: 1333–13336. 6 Alger SA, Malone M, Cerulli J, Fein S, Howard L. Beneficial effects of pharmacotherapy on weight loss, depressive symptoms, and eating patterns in obese binge eaters and non-binge eaters. Obes Res 1999; 7 : 469–476. 7 Katz DA, Maloney MJ, Sutkamp JC, McConville BJ. Lower dosages of phentermine–fenfluramine given in the afternoon: five cases with significant weight loss. Int J Eat Disord 1999; 25: 469–474. 8 Kolotkin RL, Crosby RD, Williams GR, Hartley GG, Nicol S. The relationship between health-related quality of life and weight loss. Obes Res 2001; 9: 564–571. 9 Greenway FL, Ryan DH, Bray GA, Rood JC, Tucker EW, Smith SR. Pharmaceutical cost savings of treating obesity with weight loss medications. Obes Res 1999; 7 : 523–531. 10 Steel JM, Munro JF, Duncan LJ. A comparative trial of different regimens of fenfluramine and phentermine in obesity. Practitioner 1973; 211: 232–236. 11 Hudson KD. The anorectic and hypotensive effect of fenfluramine in obesity. JR Coll Gen Pract 1977; 27: 497–501. 12 Guy-Grand B, Apfelbaum M, Crepaldi G, Gries A, Lefebvre P, Turner P. International trial of long-term dexfenfluramine in obesity. Lancet 1989; 2: 1142–1145. 13 Breum L, Astrup A, Andersen T, Lammert O, Nielsen E, Garby L et al. The effect of long-term dexfenfluramine treatment on 24-hr energy expenditure in man. A double-blind placebo controlled study. Int J Obesity 1990; 14: 613–621. 14 Mathus-Vliegen EMH, Van De Voore K, Kok AME, Res AMA. Dexfenfluramine in the treatment of severe obesity: a placebocontrolled investigation of the effects on weight loss, cardiovascular risk factors, food intake, and eating behaviour. J Intern Med 1992; 232: 119–127. 15 Bray GA, Ryan DH, Gordon D, Heidingsfelder S, Cerise F, Wilson K. A double-blind randomized placebo-controlled trial of sibutramine. Obes Res 1996; 4: 263–270. 16 Lean MEJ. Sibutramine – a review of clinical efficacy. Int J Obesity 1997; 21: S30–S36. 17 Ryan DH, Kaiser P, Bray GA. Sibutramine: a novel new agent for obesity treatment. Obes Res 1995; 3 (Suppl 4): 553S–559S. 18 James WP, Avenell A, Broom J, Whitehead J. A one year trial to assess the value of orlistat in the management of obesity. Int J Obesity Relat Metab Disord 1997; 21 (Suppl 3): S24–S30. 19 Davidson MH, Hauptman J, DiGirolamo M, Foreyt JP, Halsted CH, Heber D et al. Weight control and risk factor reduction in obese subjects treated for 2 years with orlistat: a randomized controlled trial. JAMA 1999; 281: 235–242. International Journal of Obesity 20 Connolly HM, Crary JL, McGoon MD, Hensrud DD, Edwards BS, Edwards WD et al. Valvular heart disease associated with fenfluramine–phentermine. N Engl J Med 1997; 337: 581–588. 21 Graham DJ, Green L. Further cases of valvular heart disease associated with fenfluramine–phentermine. N Engl J Med 1997; 337: 635. 22 Kurz X, Van Ermen A. Valvular heart disease associated with fenfluramine–phentermine. N Engl J Med 1997; 337: 1772–1773. 23 Cannistra LB, Davis SM, Bauman AG. Valvular heart disease associated with dexfenfluramine. N Engl J Med 1997; 337: 636. 24 Weissman NJ, Tighe Jr JF, Gottdiener JS, Gwynne JT. An assessment of heart-valve abnormalities in obese patients taking dexfenfluramine, sustained-release dexfenfluramine, or placebo. Sustained-release dexfenfluramine study group. N Engl J Med 1998; 339: 725–732. 25 Griffen L, Anchors M. Asymptomatic mitral and aortic valve disease is seen in half of the patients taking ‘phen–fen’. Arch Intern Med 1998; 158: 102. 26 Khan MA, Herzog CA, St Peter JV, Hartley GG, Madlon-Kay R, Dick CD et al. The prevalence of cardiac valvular insufficiency assessed by transthoracic echocardiography in obese patients treated with appetite-suppressant drugs. N Engl J Med 1998; 339: 713–718. 27 Wee CC, Phillips RS, Aurigemma G, Erban S, Kriegel G, Riley M et al. Risk for valvular heart disease among users of fenfluramine and dexfenfluramine who underwent echocardiography before use of medication. Annals Intern Med 1998; 129: 870–874. 28 Ryan DH, Bray GA, Helmcke F, Sander G, Volaufova J, Greenway F et al. Serial echocardiographic and clinical evaluation of valvular regurgitation before, during, and after treatment with fenfluramine or dexfenfluramine and mazindol or phentermine. Obes Res 1999; 7: 313–322. 29 Li R, Serdula MK, Williamson DF, Bowman BA, Graham DJ, Green L. Dose-effect of fenfluramine use on the severity of valvular heart disease among fen–phen patients with valvulopathy. Int J Obes Relat Metab Disord 1999; 23: 926–928. 30 Kancherla MK, Salti HI, Mulderink TA, Parker M, Bonow RO, Mehlman DJ. Echocardiographic prevalence of mitral and/or aortic regurgitation in patients exposed to either fenfluramine– phentermine combination or to dexfenfluramine. Am J Cardiol 1999; 84: 1335–1338. 31 Burger AJ, Sherman HB, Charlamb MJ, Kim J, Asinas LA, Flickner SR et al. Low prevalence of valvular heart disease in 226 phentermine–fenfluramine protocol subjects prospectively followed for up to 30 months. J Am Coll Cardiol 1999; 34: 1153–1158. 32 Teramae CY, Connolly HM, Grogan M, Miller Jr FA. Diet drugrelated cardiac valve disease: the Mayo Clinic echocardiographic laboratory experience. Mayo Clin Proc 2000; 75: 456–461. 33 Jick H, Vasilakis C, Weinrauch LA, Meier CR, Jick SS, Derby LE. A population-based study of appetite-suppressant drugs and the risk of cardiac-valve regurgitation. N Engl J Med 1998; 339: 719–724. 34 Rowland NE, Marshall M, Roth JD. Comparison of either norepinephrine-uptake inhibitors or phentermine combined with serotonergic agents on food intake in rats. Psychopharmacology (Berl) 2000; 149: 77–83. 35 Anchors M. Safer Than Phen–Fen. Prima Publishing: Rocklin, CA, USA, 1997. 36 Devlin MJ, Goldfein JA, Carino JS, Wolk SL. Open treatment of overweight binge eaters with phentermine and fluoxetine as an adjunct to cognitive-behavioral therapy. Int J Eat Disord 2000; 28: 325–332. 37 Reeve HL, Nelson DP, Archer SL, Weir EK. Effects of fluoxetine, phentermine, and venlafaxine on pulmonary arterial pressure and electrophysiology. Am J Physiol 1999; 276 (2 Part 1): L213–L219. 38 Bostwick JM, Brown TM. A toxic reaction from combining fluoxetine and phentermine. J Clin Psychopharmacol 1996; 16: 189–190. Combinations of obesity drugs LD Whigham et al 857 39 Griffen L, Anchors M. The ‘phen-pro’ diet drug combination is not associated with valvular heart disease. Arch Intern Med 1998; 158: 1278–1279. 40 Atkinson RL, Russ CS, Ciavarella PA, Owsley ES, Bibbs ML. A comprehensive approach to outpatient obesity management. J Am Diet Assoc 1984; 84: 439–444. 41 Centers for Disease Control and Prevention (CDC). Cardiac valvulopathy associated with exposure to fenfluramine or dexfenfluramine: US Department of Health and Human Services interim public health recommendations, November, 1997. MMWR Morb Mortal Wkly Rep 1997; 46: 1061–1066. 42 Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults. The evidence report. National Institutes of Health. Obes Res 1998; 6 (Suppl 2): 51S–209S. 43 Wadden TA, Berkowitz RI, Womble LG, Sarwer DB, Arnold ME, Steinberg CM. Effects of sibutramine plus orlistat in obese women following 1 year of treatment by sibutramine alone: a placebocontrolled trial. Obes Res 2000; 8: 431–437. 44 Bowen RL, Atkinson RL. Addition of orlistat to long term phentermine treatment for obesity. Obes Res 2000; 8: 118. 45 Goldstein DJ, Rampey Jr AH, Dornseif BE, Levine LR, Potvin JH, Fludzinski LA. Fluoxetine: a randomized clinical trial in the maintenance of weight loss. Obes Res 1993; 1: 92–98. 46 Goldstein DJ, Rampey Jr AH, Roback PJ, Wilson MG, Hamilton SH, Sayler ME et al. Efficacy and safety of long-term fluoxetine treatment of obesity – maximizing success. Obes Res 1995; 3 (Suppl 4): 481S–490S. 47 Sussman N, Ginsberg DL, Bikoff J. Effects of nefazodone on body weight: a pooled analysis of selective serotonin reuptake 48 49 50 51 52 53 54 inhibitor- and imipramine-controlled trials. J Clin Psychiatry 2001; 62: 256–260. Hansen D, Astrup A, Toubro S, Finer N, Kopelman P, Hilsted J, et al., Goulder For The STORM Study Group M. Predictors of weight loss and maintenance during 2 years of treatment by sibutramine in obesity Results from the European multi-centre STORM trial. sibutramine trial of obesity reduction and maintenance. Int J Obes Relat Metab Disord 2001; 25: 496–501. Wadden TA, Berkowitz RI, Sarwer DB, Prus-Wisniewski R, Steinberg C. Benefits of lifestyle modification in the pharmacologic treatment of obesity: a randomized trial. Arch Intern Med 2001; 161: 218–227. Foster GD, Wadden TA, Vogt RA, Brewer G. What is a reasonable weight loss? Patients’ expectations and evaluations of obesity treatment outcomes. J Consult Clin Psychol 1997; 65: 79–85. Rothman RB, Baumann MH, Savage JE, Rauser L, McBride A, Hufeisen SJ et al. Evidence for possible involvement of 5-HT(2B) receptors in the cardiac valvulopathy associated with fenfluramine and other serotonergic medications. Circulation 2000; 102: 2836–2841. Abenhaim L, Moride Y, Brenot F, Rich S, Benichou J, Kurz X et al. Appetite-suppressant drugs and the risk of primary pulmonary hypertension. N Engl J Med 1996; 335: 609–616. Rich S, Rubin L, Walker AM, Schneeweiss S, Abenhaim L. Anorexigens and pulmonary hypertension in the United States: results from the surveillance of North American pulmonary hypertension. Chest 2000; 117: 870–874. Ulus IH, Maher TJ, Wurtman RJ. Characterization of phentermine and related compounds as monoamine oxidase (MAO) inhibitors. Biochem Pharmacol 2000; 59: 1611–1621. International Journal of Obesity