Summary
Background
Fenfluramine and its derivatives have been associated with significant risk of developing valvular heart disease but its exact prevalence and severity are still debated.
Aim
To evaluate the clinical and echocardiographic characteristics of patients hospitalized in a cardiology centre and who had past exposure to these drugs.
Methods
Between July 2011 and February 2012, patients admitted to the hospitalization and intensive care units at the University Centre of Montpellier, France were questioned about past exposure to fenfluramine or its derivatives. In patients who reported exposure, a questionnaire assessing prescribing patterns and medical history was proposed and echocardiography performed. All of the usual echocardiographic variables were analysed. We applied criteria from a French multicentre registry for diagnosis of drug-induced valvulopathy: leaflets and subvalvular apparatus thickening and retraction, leaflets loss of coaptation, no calcification, and no stenosis.
Results
Ninety-five patients exposed to these drugs were included. The majority were female ( n = 62, 65.3%), 53.2% ( n = 50) had diabetes and 90.5% ( n = 86) were exposed to benfluorex. Mean treatment duration was 52.3 months (95% confidence interval [CI] 39.0–65.6). Valvular regurgitations were observed in 64.0% of patients ( n = 57) while 19.8% ( n = 17) had pulmonary hypertension. Highly probable fenfluramine-induced regurgitations were present in 18.6% ( n = 16) of patients, possibly fenfluramine-induced regurgitations in 38.2% ( n = 34) of patients, and unlikely fenfluramine-induced regurgitations in 25.8% ( n = 23) of patients. Highly probable fenfluramine-induced regurgitations were mild to moderate in severity in all except three patients.
Conclusion
Considering the frequency of probable or possible fenfluramine-induced regurgitations and in the absence of definite knowledge about the evolution of drug-induced valvular disease, systematic questioning about fenfluramine use may be advisable in hospitalized cardiac patients.
Résumé
Contexte
La consommation de dérivés fenfluraminiques a été associée à un sur-risque de survenue de lésions valvulaires mais la prévalence et la sévérité de ces valvulopathies est encore discutée.
Objectif
Évaluer les caractéristiques cliniques et échocardiographiques des patients hospitalisés en cardiologie et qui ont été exposés aux dérivés fenfluraminiques.
Méthodes
Entre juillet 2011 et février 2012, les patients hospitalisés en cardiologie au centre universitaire de Montpellier (France) ont été interrogés quant à une exposition à un dérivé fenfluraminique. Un questionnaire était remis aux patients exposés afin de préciser les modalités de prescription et les antécédents médicaux. Une échocardiographie était réalisée et nous avons utilisé les critères du Registre multicentrique français pour le diagnostic de lésions valvulaires d’origine toxique liées à l’exposition aux dérivés fenfluraminiques : épaississement et rétraction de la valve et de l’appareil sous-valvulaire, défaut de coaptation, absence de calcification, et absence de sténose.
Résultats
Quatre-vingt-quinze patients exposés aux dérivés fenfluraminiques ont été inclus. La majorité d’entre eux étaient des femmes ( n = 62, 65,3 %) ; 53,2 % ( n = 50) avaient un diabète sucré et 90,5 % ( n = 86) avaient été exposés au benfluorex. La durée moyenne de traitement était de 52,3 mois (IC 95 % 39,0–65,6). Des fuites valvulaires ont été observées chez 64 % des patients ( n = 57), tandis que 19,8 % ( n = 17) avaient une hypertension artérielle pulmonaire. Les fuites valvulaires probablement imputables aux dérivés fenfluraminiques concernaient 18,6 % ( n = 16) des patients, les fuites valvulaires possiblement imputables 38,2 % ( n = 34) et les fuites valvulaires non imputables 25,8 % ( n = 23). Les fuites valvulaires probablement imputables étaient toutes de sévérité légère à modérée exceptée pour 3 cas.
Conclusions
Étant donné la fréquence des fuites valvulaires probablement ou possiblement imputables aux dérivés fenfluraminiques et en l’absence de certitude quant au potentiel évotutif des lésions valvulaires d’origine toxique, la recherche systématique d’une exposition aux dérivés fenfluraminiques pourrait être envisagée en pratique quotidienne pour les patients hospitalisés en cardiologie.
Background
Fenfluramine and its derivatives are sympathomimetic amines with an anorectic action mediated through the activation of serotoninergic pathways in the brain. These drugs have been suspected to cause valvular heart disease (VHD) and pulmonary artery hypertension (PAH) . Fenfluramine was introduced first in France in 1963 under the brand name Ponderal ® ; then as the dextrorotatory isomer dexfenfluramine in 1985 under the name Isomeride ® ; and as benfluorex in 1976 under the name Mediator ® . Ponderal ® and Isomeride ® were withdrawn from the market in 1997, followed by Mediator ® in 2009, after the publication of case reports reporting cardiovascular adverse drug reactions.
In 2010 Frachon et al. established an association between benfluorex and unexplained mitral regurgitation. Exposure to benfluorex was found in 70% of patients with mitral insufficiency of undetermined cause (cases) and in 5.6% of the control group. Benfluorex was associated with a high-risk of mitral regurgitation (odds ratio [OR] 17.1, 95% confidence interval [CI] 3.5–83, P < 0.01). In the REGULATE trial , Derumeaux et al. reported that minimal valvular abnormalities were detected in 51% of patients before any exposure to benfluorex. In the Health Insurance Fund for Salaried Workers (CNAMTS) survey , the adjusted relative risk of hospitalization for valvular regurgitation was 3.1 times greater (95% CI 2.4–4.1) in patients exposed to benfluorex and the risk of hospitalization for heart valve replacement was multiplied by 3.9 (95% CI 2.6–6.1). The risk of developing mitral and aortic regurgitations was also higher with benfluorex exposure (adjusted relative risk 2.5, 95% CI 1.9–3.7 and 4.4, 95% CI 3.0–6.6, respectively). According to Hill et al. , benfluorex was responsible for 465 deaths after a follow-up of 5.5 years, and for 3500 hospitalizations for valvular insufficiency and 1750 cardiac surgeries after 4 years of follow-up. Valvular heart disease related to benfluorex mainly affects the mitral and/or aortic valves, but can simultaneously alter the four valves .
The aim of this single-centre prospective study was to report the clinical characteristics and echocardiographic features of patients admitted to our cardiology department and who had a history of treatment with fenfluramine or its derivatives.
Methods
Subjects
Between July 2011 and February 2012, all patients admitted to the hospitalization unit and the intensive care unit of the University Hospital of Montpellier were questioned about past exposure to fenfluramine or its derivatives, dexfenfluramine and benfluorex.
Clinical data
Patients who reported previous exposure to fenfluramine or its derivatives were invited to complete a questionnaire that assessed prescribing patterns, medical history and symptoms. Patients with advanced dementia were excluded from the study. The data from questionnaires were validated against any medical reports available on the hospital medical database, to correct any error or approximation.
Ischaemic heart disease was defined as history of acute coronary syndromes and/or the presence of significant coronary lesions. History of heart failure was defined as previous hospitalization for acute pulmonary oedema or right heart failure or significant exertional dyspnoea (New York Heart Association class II). Chronic respiratory disease was defined as chronic restrictive or obstructive pulmonary disease. We searched for evidence of a primary or secondary cause of VHD, including valvular prolapse, active or previous infectious endocarditis, rheumatic heart disease or history of rheumatic fever, radiation-induced VHD, congenital disease, associated connective disease and/or vasculitis, ischaemic or ventricular dilatation-associated regurgitation. We identified patients with a history of valve surgery and gathered the type of surgery performed and the pattern of valvular lesions.
Doppler echographic data
Echocardiography was performed for each patient who reported exposure to fenfluramine or its derivatives. All of the usual echocardiographic parameters were analysed, with special attention to valve aspects and pulmonary pressure evaluation. We applied the criteria for the diagnosis of drug-induced VHD identified in a French multicentre registry , which included: leaflets and subvalvular apparatus thickening and retraction, leaflets loss of coaptation, no calcification and no stenosis. Before the diagnosis of highly probable drug-induced valvulopathy could be made, other causes of restrictive VHD were excluded, in particular remodelling of the left ventricle and rheumatic heart disease. We retained the following criteria as indicating highly probable drug-induced valvulopathy: pure regurgitation, no calcification, leaflets retraction without significant left ventricular dysfunction in the case of mitral regurgitation, no dilatation of the ascending aorta in the case of aortic regurgitation. We thought that thickening and reduced mobility of leaflets, and thickening of subvalvular apparatus were more difficult to establish. We retained the following criteria as indicating unlikely drug-induced valvulopathy: stenosis, calcifications, and commissural fusions in the case of mitral regurgitation. Mitral and aortic regurgitations, which did not match with criteria for either highly probable or unlikely drug-induced valvulopathy, were considered as possible drug-induced valvulopathies. PAH was considered significant when > 40 mmHg. Patients who had history of valve surgery were included only if the valvulopathy was discovered ≥ 3 months after the start of the drug.
Several options were considered for obtaining echocardiographic data. For patients admitted to the intensive care unit, transthoracic echocardiography images taken on admission were recovered. Patients admitted to the hospitalization unit were referred for echocardiographic examination at the echocardiography laboratory. If an echocardiogram could not be performed, we gathered the echocardiographic data available on the hospital internal network. In the absence of any such information, we contacted the out-of-hospital cardiologist to obtain any echocardiographic data, as accurately as possible in the case of valvular lesions. The echocardiography equipment used included: Aloka, Philips iE33, General Electric VIVID7 and General Electric VIVID9.
All of the available echocardiographic information was analysed, irrespective of when the scans were performed. Transesophageal echocardiography images, when available, were preferred for the analysis of valves. In patients with a history of valve surgery, we sought to investigate the type of VHD at the time of surgery. Echocardiographic evaluation criteria were those established by the American Society of Echocardiography .
Statistical analysis
Clinical and echocardiographic characteristics were expressed as mean and 95% CI for continuous variables and as counts and percentages for non-continuous variables.
Results
Study population and clinical data
The study involved 95 patients who completed the questionnaires ( Table 1 ). There was predominance of women (65.3%) and the mean age was 66.8 (95% CI 64.2–69.3) years. Most patients were overweight, with a mean body mass index (BMI) of 29.5 (95% CI 28.3–30.7) kg/m 2 . The most common comorbidities were hypertension (72.3%), dyslipidemia (68.1%), diabetes mellitus (53.2%) and coronary artery disease (49.5%). Hospitalization was linked to the management of coronary artery disease in 47.3% of patients, heart failure in 18.7% and VHD in 8.8%.
| Patients ( n = 95) | |
|---|---|
| Demographic data | |
| Women | 62/95 (65.3) |
| Age (years) | 66.8 ± 2.5 |
| BMI (kg/m 2 ) | 29.5 ± 1.2 |
| Cardiovascular risk factors | |
| Past or current smoker | 43/95 (45.7) |
| Hypertension | 68/95 (72.3) |
| Dyslipidaemia | 64/95 (68.1) |
| Diabetes mellitus | 50/95 (53.2) |
| Clinical data | |
| Coronary artery disease | 47/95 (49.5) |
| Atrial fibrillation | 22/95 (23.2) |
| Heart failure | 25/95 (26.3) |
| Lung disease | 19/95 (20.2) |
| Chronic renal failure | 24/95 (25.8) |
| Rheumatic fever | 3/95 (3.3) |
| Infectious endocarditis | 2/95 (2.1) |
| Autoimmune disease | 6/95 (6.4) |
| Radiotherapy | 6/95 (6.5) |
| Neoplasia | 20/95 (21.5) |
| Reason for admission | |
| Acute coronary syndrome | 14/95 (15.4) |
| Angina | 12/95 (13.2) |
| Positive stress test | 15/95 (16.5) |
| Coronary angioplasty | 2/95 (2.2) |
| Heart failure | 17/95 (18.7) |
| Assessment of valvulopathy | 8/95 (8.8) |
| Pulmonary artery hypertension | 5/95 (5.5) |
| Infectious endocarditis | 1/95 (1.1) |
| Atrial fibrillation | 5/95 (5.5) |
| Symptoms | |
| Dyspnoea | 68/95 (71.6) |
| Chest pain | 56/95 (58.9) |
| Palpitations | 36/95 (37.9) |
| History of valve surgery | 17/95 (18.7) |
| Valve intervention | |
| Mitral valve replacement | 3/17 (17.6) |
| Mitral valve repair | 2/17 (11.8) |
| Mitral valve dilatation | 1/17 (5.9) |
| Aortic valve replacement | 8/17 (47.1) |
| Aortic valve dilatation | 2/17 (11.8) |
| TAVI | 4/17 (23.5) |
| Tricuspid valve repair | 1/17 (5.9) |
| Valve disease | |
| Mitral regurgitation | 3/17 (17.6) |
| Mitral stenosis | 1/17 (5.9) |
| Aortic regurgitation | 3/17 (17.6) |
| Aortic stenosis | 10/17 (58.8) |
Thirty-two patients (34.0%) had known VHD; 5 patients (5.3%) had heart disease officially linked to fenfluramine or a derivative (PAH in 4 patients and aortic regurgitation in 1 patient). Sixty-eight patients (71.6%) reported dyspnoea, 58.9% chest pain and 37.9% palpitations. Seventeen patients (18.7%) reported surgical or percutaneous valve interventions that had already been completed or were scheduled; the distribution of different types of surgery is summarized in Table 1 . The most common intervention was surgical aortic valve replacement and the most common heart valve disease was aortic stenosis. Two of the three cases of aortic regurgitation were caused by infective endocarditis. Excluding infective endocarditis, surgical or percutaneous valve interventions for mitral or aortic regurgitations were reported in 3 of 95 (3.2%) patients.
Treatment
Of the 95 patients, 90.5% had been treated with benfluorex, 35.8% with dexfenfluramine and 1.1% with fenfluramine ( Table 2 ). Twenty-five patients (26.3%) had been exposed to both benfluorex and dexfenfluramine. The primary indication for treatment was overweight, followed by diabetes mellitus; misuse of benfluorex was reported in a large proportion of the population (47.7%). Treatment was initiated after 1997 (date of withdrawal of dexfenfluramine) in 36.0% of patients. The mean duration of treatment was 52.3 (95% CI 39.0–65.6) months, and 90.5% had been exposed for ≥ 3 months and 31.6% for ≥ 5 years.
