The Progression of Fenfluramine-Associated Valvular Heart Disease Assessed by Echocardiography
Steven T. Mast,J. Jollis,T. Ryan,K. Anstrom,J. Crary
DOI: https://doi.org/10.7326/0003-4819-134-4-200102200-00008
IF: 39.2
2001-02-20
Annals of Internal Medicine
Abstract:In 1997, the first report of significant valvular heart disease occurring in patients exposed to the dietary suppressant fenfluramine was published (1). The investigators described 24 patients from North Dakota and Minnesota who took combination therapy with fenfluramine and phentermine for an average of 11 months and developed clinically significant left- and right-sided valvular regurgitation. In many cases, the mitral valve leaflets, especially the posterior leaflet, appeared restricted and thickened on echocardiography. In 1997, the U.S. Food and Drug Administration described results from five centers that did echocardiography in patients taking fenfluramine and dexfenfluramine and found that nearly one third of exposed patients met their case definition of at least moderate mitral regurgitation or mild aortic regurgitation (2). Furthermore, the prevalence of significant regurgitation was higher in patients who took these drugs for longer than 6 months. As a result, fenfluramine and dexfenfluramine were voluntarily withdrawn from the U.S. market in September 1997. Several reports have described varying prevalences of valvular heart disease in patients exposed to fenfluramines, but information on the natural history of fenfluramine-associated valvular heart disease is limited (3-7). The likelihood of regression, stabilization, or progression of valvular regurgitation after exposure to fenflura-mine remains largely unknown. Such information is needed to counsel patients and guide medical or surgical treatment in patients with clinically significant disease. In a small study of five patients exposed to fenfluraminephentermine who developed clinically significant valvular regurgitation, three had improvement in the degree of regurgitation after discontinuation of therapy (8). In addition, follow-up echocardiography in 1073 patients exposed to dexfenfluramine or sustained-release dexfenfluramine for a short time suggested improvement in the relatively mild valvular regurgitation observed on the initial postexposure echocardiogram (9). In a continuation and expansion of the original case series of patients from North Dakota and Minnesota, we compared echocardiograms obtained after cessation of fenfluramine therapy with subsequent echocardiograms to further describe the progression of fenfluramine-associated valvular disease. Methods Patient Selection The institutional review board at MeritCare Medical Center, Fargo, North Dakota, approved the study. Using the echocardiography database at MeritCare Medical Center, we identified all patients who had potential anorectic exposure according to the indication for the procedure and who had had two or more echocardiograms. Type and duration of drug exposure were subsequently confirmed by medical record review and patient contact. We initially identified all 69 patients who had taken fenfluramine or dexfenfluramine and had had serial echocardiography as part of routine care at least 100 days apart between December 1994 and February 1999. Of these patients, we excluded 17 in whom the initial echocardiography was done before treatment with a fenfluramine and 2 patients who did not have at least mild regurgitation on initial echocardiogram. Clinical data, including type of dietary suppressant taken and duration of therapy, presence of cardiac symptoms, preexisting cardiac disease, other medications taken, and need for cardiac surgery, were extracted from the medical record or obtained through patient interview. Echocardiography Complete echocardiograms were obtained as part of routine clinical practice. Before June 1997, these studies were obtained to evaluate cardiac symptoms alone in 3 patients, heart murmurs alone in 6 patients, and both symptoms and a murmur in 2 patients and for unspecified reasons in 5 patients. After June 1997, when the possibility of fenfluramine-associated valve regurgitation was more clearly recognized, 14 patients without symptoms or murmurs were studied to check for this disease. Two patients with symptoms alone, 14 patients with murmurs alone, and 4 patients with both symptoms and murmurs were also studied after this date. Echocardiograms were acquired by using Acuson (Mountain View, California) and Hewlett-Packard (Andover, Massachusetts) ultrasonographic imaging systems. When feasible, images were obtained in the parasternal, apical, and subcostal positions, and all cardiac valves were interrogated by using spectral, continuous wave, and color Doppler in accordance with American Society of Echocardiography standards (10). When multiple echocardiograms were available, only the first and last echocardiograms in the series were included in the analysis. Serial echocardiograms were reviewed on videotape by two echocardiographers blinded to order of image acquisition. Severity of valvular regurgitation and the presence of valvular restriction were assessed. Differences in interpretation were settled by consensus of the two readers after concurrent review of the echocardiograms. Valvular Abnormalities Valvular motion was described as normal when a given leaflet displayed full mobility, mildly restricted when leaflet motion was restricted but was not impaired by more than 50% of normal excursion, moderately restricted when leaflet motion was impaired by more than 50% of normal excursion, and severely restricted when the leaflet was completely immobile. Normal terminal excursion points for the aortic, anterior mitral valve, and posterior mitral valve leaflets were the aortic wall, left ventricular outflow tract, and posterior left ventricular wall, respectively. Severity of valvular regurgitation was classified as trivial, mild, moderate, or severe by using color-flow Doppler according to accepted definitions (10). Valvular regurgitation was considered indeterminate when the valve was not interrogated or could not be clearly seen. Statistical Analysis Contingency tables were constructed to compare regurgitation and leaflet mobility on initial and follow-up echocardiograms. Paired tests for marginal homogeneity using exact tests were applied to assess the change in severity of regurgitation or leaflet restriction (StatXact software, version 4.01, Cytel Software Corp., Cambridge, Massachusetts). In additional comparisons of patient and echocardiographic characteristics, we used unpaired t-tests to compare continuous descriptors and chi-square analyses to compare categorical descriptors. Results Clinical Information Paired serial echocardiograms were reviewed for 50 patients. Patient characteristics, including age, sex, body weight, and coexisting illness, were similar to those in previous nationally representative studies (Table 1) (6, 7). Of these patients, 39 had taken fenfluramine plus phentermine; 1 had taken fenfluramine alone; 1 had taken dexfenfluramine alone; and 9 had taken various combinations of fenfluramine, phentermine, and dexfenfluramine. The mean (SD) duration of drug treatment was 447 311 days (range, 65 to 1432 days). The mean time between cessation of dietary suppressants and obtaining the first echocardiogram was 190 214 days (range, 1 to 981 days), and the mean time between the first and last echocardiograms was 356 224 days (range, 100 to 1454 days). Table 1. Patient Characteristics after Exposure to Drug Cardiac symptoms, including dyspnea, chest pain, palpitations, and fatigue, were present in a substantial number of patients after drug exposure (Table 1). The most common symptoms at baseline were dyspnea and fatigue. Of 15 patients with these symptoms at initial echocardiography for whom follow-up data were available, 2 had complete resolution at follow-up. Of 21 patients without such symptoms at initial echocardiography, only 1 patient had symptoms at follow-up. This patient developed dyspnea and fatigue after myocardial infarction. One patient underwent mitral valve replacement and 1 patient underwent bypass surgery without valve surgery. One patient died of complications of systemic lupus erythematosus and renal transplantation. Cardiac murmurs were noted in 27 patients. Preexisting acquired or congenital heart disease was rare in this cohort; 2 patients had a history of coronary disease, 3 patients had previous rheumatic heart disease, and 1 patient had congenital heart disease. Echocardiographic Follow-up Table 2 shows the degree of change in regurgitation between the initial postexposure echocardiogram and the follow-up echocardiogram. For mitral and aortic regurgitation, most patients were unchanged, while a large number improved by one or more grades. For tricuspid and pulmonic regurgitation, most patients improved. In all valves, worsening of regurgitation was rare. Table 2. Changes in Valvular Regurgitation on Serial Echocardiography Changes in severity of regurgitation and leaflet restriction in individual valves are shown in Table 3. Significant mitral regurgitation was commonly seen in the initial postexposure echocardiogram; 38 patients (76%) had at least mild mitral regurgitation and 16 patients had severe mitral regurgitation. Mitral regurgitation improved by at least one grade in 17 patients (P=0.001) and aortic regurgitation improved by at least one grade in 19 patients (P=0.004). Mitral regurgitation remained unchanged in 19 patients, improved in 17 patients, and worsened in 2 patients (P=0.001). When mitral regurgitation improved, it was typically by one grade of severity; 3 patients had two or more grades of improvement. Table 3. Changes in Valvular Regurgitation on Serial Echocardiography Thirty patients had posterior mitral valve leaflet restriction on initial study (Table 4). The degree of restriction was unchanged in 36 patients, improved in 11 patients, and worsened in 2 patients. Restriction of the anterior mitral valve leaflet was rare, occurring in only 4 patients on initial or follow-up echocardiography. Table 4. Changes in Mitral Posterior Leaflet Restriction on Serial Echocardiography Aortic reg