[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
Purchase Options:
[Skip to Content Landing]
April 5, 2000

Heart Valve Disorders and Appetite-Suppressant Drugs

Author Affiliations

Author Affiliation: Department of Medicine, Boston University School of Medicine, Lexington, Mass.

JAMA. 2000;283(13):1738-1740. doi:10.1001/jama.283.13.1738

An association between appetite-suppressant medications and cardiac valve disorders is now generally believed to have been established, but remains an important scientific and clinical issue. Even though the 2 most commonly implicated agents, fenfluramine and dexfenfluramine, were withdrawn from the US market in 1997 and despite the recent multibillion dollar class action settlement by the manufacturer of these drugs,1 ongoing research continues to attempt to more accurately and more completely characterize the pathophysiology and natural history of anorexigen-associated valvular heart disorders. However, the totality of the evidence to date favoring a causal connection between fenfluramines and cardiac valve disorders is persuasive, if somewhat complex.

In the initial 1997 report of cardiac valvular abnormalities associated with appetite-suppressant drugs, Connolly et al2 described 24 patients who had developed symptomatic heart valve disorders in the absence of preexisting conditions after receiving phentermine and fenfluramine. The patients had moderate to severe symptomatic aortic regurgitation (AR), mitral regurgitation (MR), or both, and had taken both drugs for at least 4 months. Five of the patients required valve replacement. The valves had an unusual morphologic appearance consisting of glistening white leaflets and chordae with diffuse thickening and were considered "identical" to those seen in carcinoid or ergotamine-induced valve disease. Carcinoid and ergotamines are associated with elevated levels of serotonin, and because fenfluramines stimulate the secretion of serotonin, it was suggested that the mechanism by which fenfluramines may cause heart valve disorders could be related to elevated levels of serotonin.

At the same time, the Food and Drug Administration described the case histories of 36 recipients of fenfluramines with similar heart valve disorders.3 Six of the patients had received fenfluramine or dexfenfluramine alone. Subsequently, the Food and Drug Administration reported the result of echocardiographic studies on 271 patients that found that on average 32% had valve abnormalities after receiving fenfluramines.4 As a result of these observations, fenfluramines were voluntarily withdrawn from the market.

These reports prompted 2 types of studies on the relation of fenfluramines and heart valve disorders. The first involved echocardiographic examination of asymptomatic obese patients who had received fenfluramines. The second type was a formal long-term, population-based follow-up study of a large number of fenfluramine users designed to identify patients who had developed clinically diagnosed symptomatic heart valve disorders.

In a study of 496 otherwise healthy obese subjects, Khan et al5 reported a prevalence of heart valve abnormalities of 22.7% in recipients of fenfluramines compared with 1.3% in comparable nonrecipients. The primary valve abnormality noted was AR, most often considered to be trace to mild. The duration of use of fenfluramines varied, but no details were provided. Weissman et al6 reported the results of a placebo-controlled randomized study of dexfenfluramine and found a modestly elevated prevalence of AR (5.8%) in 718 dexfenfluramine recipients compared with 3.6% in 354 control subjects. Of importance, the average duration of treatment was only 71 to 72 days. Griffen and Anchors7 obtained echocardiograms (ECHOs) on 22 patients who had taken phentermine/fenfluramine for more than 3 months and reported that 10 (46%) had "significant" asymptomatic AR. By contrast, Griffen and Anchors8 also obtained ECHOs on 60 patients who had received a combination of phentermine and fluoxetine for weight control and found that only 1 patient had mild AR.8

More recently, Shively and colleagues9 described the risk of valve disorders diagnosed by ECHO in 412 subjects, among whom 223 had taken dexfenfluramine for a mean duration of 6.9 months, and 189 subjects matched on age, sex, and body mass index who were unexposed. Valvular regurgitation, most often AR, was observed in 7.6% of subjects in the dexfenfluramine-exposed group compared with 2.1% in the nonexposed group (P=.01). In a study by Burger et al,10 18 (8%) of 226 phentermine/fenfluramine users were found to have mild to moderate AR by ECHO, but no comparison group of nonobese users was reported.

Taken together, echocardiographic studies of asymptomatic obese patients have consistently found an increased prevalence of valve disorders, particularly AR, in users of fenfluramines compared with nonusers.5-8 The prevalence of valve disorders in nonexposed subjects has varied from 1% to 4%, whereas the prevalence of such disorders in fenfluramine users has been reported to be as high as 32%.

Jick and colleagues11 conducted a population-based, long-term follow-up study of more than 17,000 obese patients using a database that provides virtually complete information over 8 years about drugs prescribed, clinical diagnoses, and records on referrals and hospitalizations of more than 3 million patients in the United Kingdom. The study, which included more than 8000 recipients of fenfluramines alone and a comparable group of obese nonrecipients, identified 22 symptomatic patients referred or hospitalized with a first-time computer-recorded clinical diagnosis of a valve disorder including 11 patients with newly diagnosed valve disorders in the absence of predisposing conditions (ie, idiopathic) and 11 with predisposing conditions such as congenital or rheumatic heart disease. Among the 11 patients considered to have idiopathic valve disorders, all were recipients of fenfluramines, 9 had received the drug for more than 3 months, and a similar proportion had aortic insufficiency. No such cases of valvulopathy were found in nonrecipients. By contrast, among the 11 patients with predisposing conditions, use of fenfluramines was similar to that in the base population. This study provided evidence that heart valve disorders associated with fenfluramines can lead to clinically diagnosed symptomatic illness and also indicated that the prevalence of clinically symptomatic heart valve disorders associated with fenfluramines was low (approximately 1 per 1000 recipients) and occurred primarily in patients with aortic insufficiency who had taken fenfluramines for more than 3 months. The results suggest that the vast majority of the echocardiographic-described valve disorders in asymptomatic patients associated with fenfluramines5-10 remain asymptomatic for many years and perhaps indefinitely.

In this issue of THE JOURNAL, Gardin and colleagues12 report the results of their study of 1473 patients from 25 clinical centers in the United States. The results again demonstrate a significantly higher prevalence of asymptomatic AR, primarily mild, in 479 recipients of dexfenfluramine (8.9%) and 455 recipients of phentermine/fenfluramine (13.7%), compared with 539 (4.1%) matched controls. In a secondary analysis in which patients with echocardiographic evidence of heart valve abnormalities characteristic of other valvular pathology (eg, mitral valve prolapse) and those who had an ECHO prior to anorexigen use were excluded, the prevalence of AR also was significantly increased among anorexigen-treated patients (6.3% for dexfenfluramine, 13.9% for phentermine/fenfluramine, and 3.4% for controls). Moreover, although the authors report no significant difference in the prevalence of MR of at least moderate severity among the 3 groups, there was a significant increase in the prevalence of MR in both anorexigen-treated groups when all grades of regurgitation were evaluated, primarily due to an increase in cases of mild MR.

This study also is the first to provide considerable data on the relation of duration of use. The prevalence of AR in the subjects who had used fenfluramines for 3 months or less (n=48 for phentermine/fenfluramine; n=92 for dexfenfluramine) was 1.4%, whereas the prevalence was 13% in the 781 subjects who had used fenfluramines for more than 3 months. Furthermore, the prevalence of AR increased with increasing duration of exposure, reaching 21.2% in subjects who had received fenfluramines for more than 18 months.

However, the findings of Gardin et al should be interpreted with care. The mean time from the last dose of the anorectic agent to performance of the ECHO was 5.3 months for patients in the dexfenfluramine group and 6.8 months in the phentermine/fenfluramine group. In a recently published small prospective study in which patients who had received anorectic agents as part of a clinical trial and had ECHOs obtained at the termination of the trial and then at 6 months, Hensrud et al13 reported that the findings of anorexigen-linked valvulopathy noted at the time of the first ECHO did not progress after cessation of use of anorexigens, and in some patients appeared to improve over time. As pointed out by Asinger,14 ECHO screening several months after cessation of anorexigen treatment may result in a lower incidence of valvular abnormalities than screening performed while patients were taking the drug or shortly thereafter. Accordingly, the data reported by Gardin et al may underestimate the actual prevalence of heart valve abnormalities related to the use of these drugs.

Several factors apparent from the body of evidence linking fenfluramines and valve disorders appear to favor a causal connection. The majority of echocardiographic studies comparing asymptomatic fenfluramine recipients with comparable nonrecipients show a substantially increased prevalence of valve disorders in the fenfluramine-treated patients. The most common abnormality is AR, which is most often minor and benign (ie, associated with no cardiac symptoms). Clinically important valvular disease attributable to fenfluramines appears to be uncommon.

Moreover, and of critical importance to a causal inference, there is now persuasive evidence of a large duration effect.12 In the United Kingdom follow-up study,11 92% of patients with symptomatic valve disorders, primarily AR, had used fenfluramines for more than 3 months, whereas only 23% of subjects in the base population received fenfluramines for more than 3 months. In the series reported by Connolly et al,2 22 of 24 cases occurred in fenfluramine recipients who used the drug for more than 3 months, despite the fact that only 20% to 30% of fenfluramine users in the United States are estimated to have taken the drug for that long. The study by Gardin et al demonstrates a clear duration effect of the fenfluramine–heart valve disorder association. In addition, a recent study by Li et al15 suggests that the severity of valvulopathy is associated with use of higher doses of fenfluramines.

Millions of patients were prescribed fenfluramines prior to 1997. For the substantial majority who took the drug for less than 3 months, the risk of heart valve disorders appears to be minimal. In those who took the drug longer than 3 months, many will have developed echocardiographic evidence of cardiac valve disorders, particularly mild AR. In the majority of instances, these abnormalities most likely are benign and are unlikely to lead to clinical disease. However, a small proportion of patients have substantially increased risk for clinically important valvulopathy and cardiovascular consequences as a result of taking anorexigens. However, because fenfluramines have been unavailable since 1997, judgments about the overall consequences of fenfluramine use are likely to be limited to the results of those studies already completed.

 Nationwide class action settlement agreement with American Home Products Corporation. Available at: http://www.settlementdietdrugs.com/. Accessed March 16, 2000.
Connolly HM, Crary JL, McGoon M.  et al.  Valvular heart disease associated with fenfluramine-phentermine.  N Engl J Med.1997;337:581-588.Google Scholar
Graham DJ, Green L. Further cases of valvular heart disease associated with fenfluramine-phentermine.  N Engl J Med.1997;337:635.Google Scholar
Food and Drug Administration.  FDA analysis of cardiac valvular dysfunction with the use of appetite suppressants. Available at: http://www.fda.gov/cder/news/slides/sld001.htm. Accessed February 17, 1998.
Khan M, Herzog C, St. Peter J.  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.Google Scholar
Weissman N, Tighe J, Gottdiener J. An assessment of heart valve abnormalities in obese patients taking dexfenfluramine, sustained release dexfenfluramine or placebo.  N Engl J Med.1998;339:725-732.Google Scholar
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.Google Scholar
Griffen L, Anchors M. The "phen-pro" diet drug combination is not associated with valvular heart disease.  Arch Intern Med.1998;158:1278-1279.Google Scholar
Shively BK, Roldan CA, Gill EA.  et al.  Prevalence and determinants of valvulopathy in patients treated with dexfenfluramine.  Circulation.1999;100:2161-2167.Google Scholar
Burger AJ, Sherman HB, Charlamb MJ.  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.Google Scholar
Jick H, Vasilakis C, Weinrauch LA.  et al.  A population-based study of appetite-suppressant drugs and the risk of cardiac-valve regurgitation.  N Engl J Med.1998;339:719-724.Google Scholar
Gardin JM, Schumacher D, Constantine G, Davis KD, Leung C, Reid CL. Valvular abnormalities and cardiovascular status following exposure to dexfenfluramine or phentermine/fenfluramine.  JAMA.2000;283:1703-1709.Google Scholar
Hensrud DD, Connolly HM, Grogan M.  et al.  Echocardiographic improvement over time after cessation of use of fenfluramine and phentermine.  Mayo Clin Proc.1999;74:1191-1197.Google Scholar
Asinger RW. The Fen-Phen controversy.  Mayo Clin Proc.1999;74:1302-1304.Google Scholar
Li R, Serdula MK, Williamson DF.  et al.  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.Google Scholar