Author + information
- Received June 11, 2002
- Revision received October 17, 2002
- Accepted October 31, 2002
- Published online March 19, 2003.
- Marcel J.M Kofflard, MD†,
- Folkert J Ten Cate, MD, FACC*,* (, )
- Chris van der Lee, MD* and
- Ron T van Domburg, MSc*
- ↵*Reprint requests and correspondence:
Dr. Folkert J. Ten Cate, Thoraxcenter Ba 304, Erasmus MC, Dr. Molewaterplein 40, 3015 GD Rotterdam, The Netherlands.
Objectives This study evaluates the clinical course and identifies risk factors for sudden cardiac death (SCD) and clinical deterioration in hypertrophic cardiomyopathy (HCM) in a large community-based population. Comparison was made with data from six tertiary referral and six nonreferral institutions.
Background Hypertrophic cardiomyopathy is a disease with marked heterogeneity in clinical presentation and prognosis. Risk factors for SCD are not well defined in patients free of referral bias.
Methods Between 1970 and 1999, 225 consecutive patients (mean age [±SD] 41±16 years) were examined and followed at yearly intervals.
Results Forty-four deaths were recorded of which 27 cases were cardiovascular. Fourteen patients died suddenly, six were successfully resuscitated, and seven patients died of congestive heart failure. The annual mortality, annual cardiac mortality, and annual mortality due to sudden death were 1.3%, 0.8%, and 0.6%, respectively. At least one New York Heart Association (NYHA) functional class deterioration was reported in 33% of the patients with a significant (≥50 mm Hg) left ventricular outflow tract (LVOT) gradient in contrast to 7% without obstruction. The presence of syncope was related to SCD (p < 0.05). Younger age and more severe functional limitation distinguishes patients from in hospital-based centers from the ones in community-based centers.
Conclusions Hypertrophic cardiomyopathy is a benign disease in an unselected population with a low incidence of cardiac death. Syncope was associated with a higher incidence of SCD and patients with a significant LVOT obstruction were more susceptible to clinical deterioration.
Hypertrophic cardiomyopathy (HCM) is a complex cardiac disease with a heterogeneous genetic, morphologic, and clinical spectrum (1,2). In recent years, numerous mutations in different genes, encoding for proteins of the cardiac sarcomere, have been found to be responsible for this disease (3–6). Unfortunately, the same mutation will not always result in identical morphological and clinical features (7,8). Environmental as well as genetic factors may modify the phenotypic expression of the mutated gene (3,8,9).
The diverse clinical and genetic characteristics make it difficult to establish general guidelines for the treatment of symptoms and to predict which patient is prone to sudden death. Most of our knowledge concerning the natural history and risk for premature sudden cardiac death (SCD) has been derived from the studies of hospital-based populations (10–18). These studies reported an annual mortality rate of 2% to 4% in adults and 6% in children. However, the clinical outcome and perception of risk factors in HCM is profoundly affected by a bias in patient selection (19). Studies from nonreferral centers suggested a more benign clinical course and a fairly favorable prognosis with an annual mortality of ≤1% (20–26). However, risk factors for SCD in patients with HCM have not been extensively studied in these community-based populations. The purpose of the present study was to evaluate the clinical course and to identify the presence of risk factors for SCD and clinical deterioration in a population of relatively unselected patients with HCM. We compare our results with the results from hospital-based and community-based populations.
Before 1979, only M-mode and two-dimensional echocardiography (no Doppler) were available at our institution and the diagnosis of HCM was also made by the presence of typical clinical, electrocardiographic, and/or angiographic findings. The diagnosis of HCM was later based on the echocardiographic finding of a nondilated hypertrophic left ventricle in the absence of diseases known to cause ventricular hypertrophy (1,2). Patients who visit our HCM clinic are relatively unselected and consist of patients who are: 1) diagnosed and treated at our center; 2) referred from the community for establishment of the diagnosis or advice on management; 3) transferred from outlying hospitals for advanced care; and 4) self-referred family members of patients.
Between 1970 and 1999, 225 consecutive patients were examined initially and followed at yearly intervals. The population included 113 patients as reported earlier (22). The other 112 patients consisted of patients who were brought to our attention from 1990. At first visit, clinical characteristics including age at diagnosis of HCM, family history of HCM or sudden death in a first degree relative at a young age (<40 years) or both, symptoms, New York Heart Association (NYHA) functional class, and the drugs prescribed, were recorded. Physical examination and baseline laboratory studies were performed including electrocardiogram, M-mode and two-dimensional echocardiography (as well as Doppler echocardiography after 1985), and in the majority of patients, 24-h ambulatory Holter monitoring. Cardiac catheterization was performed only in patients with symptoms refractory to medical therapy in whom surgical treatment was considered. Follow-up information was obtained within six months from August 1999 at the outpatient clinic or, if the patient moved out of Rotterdam, from their private cardiologist. No patient was lost to follow-up. Cardiac death was defined as death caused by congestive heart failure or sudden death. All deaths witnessed within 1 h of onset of symptoms or in subjects known to be in normal functioning 24 h before, were assumed to be SCDs. Patients who were successfully reanimated after cardiac arrest during follow-up were classified as SCD for the purpose of analysis.
Review of the literature on prognosis in patients with HCM
To compare our study findings, we reviewed the literature on prognosis in HCM in both hospital-based and community-based institutions. A total of nine studies from six hospital-based centers and six studies from six community-based centers were found for comparison.
Data are expressed as mean value ± SD. Cumulative survival estimates and 95% confidence intervals (CIs) were calculated according to the Kaplan-Meier method. Among patient subgroups the log-rank test was used to evaluate differences in survival between subgroups. The yearly cumulative mortality rate was calculated on the basis of all available follow-up time. To assess risk factors for cardiac death and clinical deterioration, univariate and multivariate analysis was performed using the Cox regression model.
The chi-squared test was used to compare clinical characteristics from referral and nonreferral centers.
The clinical characteristics of the 225 patients with HCM are listed in Table 1. The majority of the patients were male (n = 130). At first visit, 63 patients (28%) were younger than 30 years and 20 (9%) were older than 65 years. The mean age (±SD) at diagnosis of HCM was 37 ± 17 years and 41 ± 16 years at presentation to our institution. A positive family history of HCM was present in 110 patients (49%), of whom 52 patients (23%) also reported a sudden death in a first degree relative. Reported symptoms included angina in 58 patients (26%), dyspnea in 81 (36%), syncope in 43 (19%), and palpitations in 43 (19%). At presentation, 100 patients (44%) were asymptomatic or had trivial symptoms (NYHA functional class I), 101 patients (45%) had mild symptoms of exertional angina and/or dyspnea (class II), and 24 patients (11%) were moderately symptomatic (class III). Initially, patients were treated with beta-blocking agents, calcium antagonists, a combination of beta-blocking agents and calcium antagonists, antiarrhythmics, or other cardiac therapy in 47 (21%), 72 (32%), 16 (7%), 18 (8%), and 11 patients (5%), respectively; 61 patients (27%) did not take medication.
On echocardiography, mean interventricular septum width was 21 ± 4 mm (16 to 40 mm), 30 patients (13%) had marked left ventricular hypertrophy (interventricular septal width ≥25 mm). Left ventricular outflow tract (LVOT) gradient, at rest or provocation was ≥50 mm Hg in 98 patients (44%), as determined by Doppler echocardiography or cardiac catheterization. At the initial visit, 7 patients (3%) presented with persistent atrial fibrillation (AF). Episodes of ventricular tachycardia (VT) on 24-h ambulatory Holter monitoring were registered in 73 of the 149 patients, in whom recordings were available. At presentation, two patients had been treated by septal myectomy for LVOT obstruction.
During follow-up, 44 deaths were reported. Twenty-seven deaths (mean age 49 ± 17 years) could be attributed to cardiovascular causes: 20 (mean age 44 ± 15 years) of these died suddenly, of whom 6 were successfully resuscitated from a witnessed cardiac arrest and 7 patients (mean age 64 ± 10 years) died of congestive heart failure (3 patients underwent orthotopic heart transplantation which is considered a heart failure death). Seventeen patients (mean age 63 ± 16 years) died of noncardiac causes: 4 patients died from cerebrovascular accidents and 13 deaths were associated with chronic obstructive lung disease (3 patients) or malignant diseases (10 patients). The annual mortality, annual cardiac mortality, and annual mortality for sudden death was 1.3%, 0.8%, and 0.6%, respectively. The 5-, 10-, and 15-year cumulative cardiac survival in our population was 96%, 91%, and 78%, respectively (Fig. 1).
During follow-up, 57 patients underwent septal myectomy, either alone (29 patients) or in combination with anterior mitral leaflet extension (22 patients) (27), mitral valve replacement (3 patients) or coronary artery bypass grafting (3 patients). There was one perioperative death. After surgical intervention, 47 patients (82%) had improvement of symptoms and 36 patients (63%) were symptom-free. A total of 10 patients had implantation of a permanent pacemaker, 3 because of postoperative atrioventricular block, 3 for sick sinus syndrome, 3 for treatment of the LVOT obstruction, and 1 patient received an internal cardiovertor defibrillator after successfully treated ventricular fibrillation.
Magnitude of left ventricular hypertrophy
Thirty patients had marked left ventricular hypertrophy at presentation. Sudden cardiac death was demonstrated in 5 of these patients (17%) compared with SCD of 15 (13%) in 195 patients without marked hypertrophy. There was no significant difference in survival for patients with or without marked left ventricular hypertrophy (p = 0.10 and p = 0.13, respectively).
Presence of LVOT obstruction
At the initial evaluation, 98 patients had a LVOT obstruction. Sudden cardiac death was established in 10 (10%) of 98 patients with obstruction and in 10 (8%) of 127 patients without obstruction. No differences in cardiac death or in SCD were found for this variable. During follow-up (mean 7.5 ± 7 years) deterioration in at least one functional NYHA class was experienced by 33 patients (33%) with a LVOT gradient and in the 57 patients who underwent septal myectomy functional deterioration was present in 24 cases (42%) (Fig. 2). In contrast, only 9 patients (7%) had functional deterioration in the nonobstructive group of 127 patients (mean follow-up 8.4 ± 6.7 years) (Fig. 3). In the patients who had surgery, the LVOT gradient did not change over time. However, “end-stage” disease with deteriorated systolic function was established in 4 of 41 the not-operated patients with an LVOT obstruction.
Presence of AF
At the latest visit, 24 patients (mean age 57 ± 11 years) had persistent AF. Persistent AF was present at initial evaluation in seven patients. Cardiac death and SCD was established in 4 (16%) and 2 (8%) patients with persistent AF compared with 23 (11%) and 18 (9%) patients in sinus rhythm (p = 0.49 and p = 0.89, respectively). The presence of AF had no influence on functional deterioration as assessed by the NYHA classification (p = 0.10). One patient with persistent AF died from an ischemic cerebrovascular accident.
Presence of VT
Episodes of nonsustained and/or sustained VT on 24-h Holter monitoring were present in 73 of 149 patients in whom one or more recordings were available. Sudden cardiac death was established in 8 (11%) patients with VT and 6 (8%) patients without VT on 24-h Holter monitoring, respectively. There was no significant difference in cardiac death and SCD in patients with VT on 24-h Holter monitoring (p = 0.76 and p = 0.78, respectively). Sustained VT was demonstrated in only seven patients of whom one patient died suddenly.
Multivariate statistical analysis
For the purpose of analysis, patients were divided according to known or suspected risk factors for SCD including age, family history of HCM and/or sudden death, history of syncope, NYHA classification, presence of AF, presence of VT on 24-h ambulatory monitoring, severe left ventricular hypertrophy (≥25 mm), and basal outflow tract obstruction (50 mm Hg). By multivariate analysis, only syncope was an independent predictor for SCD (relative risk [RR] 4.3, 95% CI 1.8 to 5.9). Furthermore, only an LVOT obstruction was an independent predictor for functional deterioration during the follow-up period (RR 1.16, 95% CI 1.08 to 1.25) for every increase of 10 mm Hg.
Review of the literature on prognosis in HCM
Nine reports originated from six referral centers and as a consequence the population in these studies consisted of highly selected patients (Table 2) (13–21). In the other six reports, patients came from six nonreferral centers and in these studies the population was considered nonselected (22–26,27)(Table 3).
From the available data it can be appreciated that both annual mortality and annual cardiac mortality is distinctly higher in referral centers compared with nonreferral centers (2.4% to 5.9% vs. 0.0% to 2.0% and 1.5% to 3.5% vs. 0.0% to 1.3%, respectively). In general, the most striking differences comparing baseline characteristics in hospital-based and community-based populations were the younger age and the more severe functional impairment of patients in referral centers (p < 0.001 and p < 0.001, respectively). This might well produce the discrepancy in clinical outcome in selected patients.
The natural history of HCM has been the subject of much debate among researchers. Observational studies from tertiary referral centers have estimated the annual cardiac death to be 2% to 4% in adults (10–17)and as high as 6% in children (13,28). These studies from major referral institutions have been influenced by selection biases (19). Indeed, in community-based studies patients with HCM have a more favorable prognosis with 1% annual mortality (20–26).
Patient-selection biases might lead to misleading conclusions about the significance of risk factors for sudden death. The present study investigates the role of risk factors for sudden death in a community-based population of patients with HCM.
Clinical characteristics in selected and unselected patients
At the time of the initial visit, mean age was 41 years and almost 90% of the patients had no or mild cardiac symptoms. These findings are comparable with two other large unselected community-based trials from Cecchi et al. (24)and Kyriadiks et al. (25)and in contrast with the findings in tertiary referral centers. This dissimilarity in clinical characteristics has undoubtedly influenced clinical outcome. However, we can only speculate how younger age and more severe functional limitation adversely influence prognosis in hospital-based patients with HCM. A malignant family history with sudden death at young age or surviving cardiac arrest and presence and symptoms of HCM in the very young have been related with adverse outcome and this might persuade the general cardiologist to refer these patients to institutions known to have special interest in HCM (1). Second, the management of patients with HCM is difficult once symptoms appear. Myocardial ischemia, abnormal diastolic function, and LVOT obstruction may all result in symptoms of angina, dyspnea, and syncope (1). The contribution of each of these mechanisms to the clinical picture differs from patient to patient. The general cardiologist may choose to refer the patient if his/her client becomes severely symptomatic (1).
Clinical outcome and risk factors for sudden death
The annual cardiac mortality in our study was 0.8%. Cardiac death is caused by heart failure or occurs sudden and unexpected. Many HCM patients who die suddenly are adolescents or young adults and have previously been symptom-free (29). The stratification of risk of sudden death remains unsettled. There seems to be agreement among investigators that there is a small subgroup of patients who are particularly at high risk for sudden death. Patients who have survived cardiac arrest, young patients with a family history of multiple SCDs, and those with a malignant gene mutation are especially at substantial risk for premature death. Young age (11,13,15), syncope (13,30), family history of HCM and SCD (13,30), functional cardiovascular status (13), presence of VT (17), positive exercise test (15,31), presence of a LVOT gradient (18), marked left ventricular hypertrophy (32), and a abnormal blood pressure response during exercise (33)are also reported as risk factors of SCD. These data are derived from hospital-based populations and, therefore, are influenced by referral bias. The latter phenomenon plays a less important role in community-based populations with HCM.
However, none of the community-based institutions have examined the role of risk factors for sudden death. In the studies by Cecchi et al. (24)and Kyriakis et al. (25)risk factors for combined sudden and heart failure-related HCM death were reported. Nevertheless, it is obvious that the process leading to death in patients with intractable congestive heart failure differs from the mechanism causing sudden unexpected death in patients with HCM. Heart failure results from systolic and/or diastolic dysfunction and SCD results from VT or ventricular fibrillation, either as a primary event or as a secondary phenomenon triggered by myocardial ischemia, diastolic dysfunction, LVOT obstruction, hypotension, or supraventricular tachycardia (1,34). In our study, risk factors for cardiac death due to congestive heart failure could not be detected. However, syncope was associated with SCD at multivariate analysis.
Syncope is a complex entity since several mechanisms may be responsible for this symptom including supraventricular and ventricular tachyarrhythmias, bradyarrhythmias, heart block, diastolic dysfunction, myocardial ischemia, autonomic dysfunction, and LVOT obstruction. Since the contribution of each of these mechanisms differs from patient to patient, the treatment of syncope must be individualized.
At multivariate analysis, only a significant LVOT gradient was related to deterioration in functional state.
Hypertrophic cardiomyopathy can be caused by a mutation in one of the genes that encode proteins of the cardiac sarcomere. At present, there are obstacles to the translation of HCM genetic research into practical clinical applications and routine clinical strategy. Exercise-induced hypotension is common in young patients with HCM and this finding has been associated with an increased risk of sudden death. However, until 1999 we did not routinely perform exercise tests in patients with HCM.
The results in our study show that prognosis of HCM in a community-based population is benign and confirm the current opinion on the natural history in HCM. The most striking finding was a substantial difference in the clinical stability of patients with and without LVOT obstruction. In nonobstructive patients, a prolonged stable clinical course is noticed. In obstructive patients, we found a progressive decline in functional status during follow-up. Furthermore, syncope was the only risk factor associated with sudden death.
- atrial fibrillation
- confidence interval
- hypertrophic cardiomyopathy
- left ventricular outflow tract
- New York Heart Association
- relative risk
- sudden cardiac death
- ventricular tachycardia
- Received June 11, 2002.
- Revision received October 17, 2002.
- Accepted October 31, 2002.
- American College of Cardiology Foundation
- Wigle E.D.
- Schwartz K.,
- Carrier L.,
- Guicheney P.,
- Komajda M.
- Hagege A.A.,
- Dubourg O.,
- Desnos M.,
- et al.
- Swan D.A.,
- Bell B.,
- Oakley C.M.,
- Goodwin J.
- Shah P.M.,
- Adelman A.G.,
- Wigle E.D.,
- et al.
- Romeo F.,
- Pelliccia F.,
- Cristofani R.,
- Martuscelli E.,
- Reale A.
- Seiler C.,
- Hess O.M.,
- Schoenbeck M.,
- et al.
- Shapiro L.M.,
- Zezulka A.
- Cannan C.R.,
- Reeder G.S.,
- Bailey K.R.,
- Melton L.J. III.,
- Gersh B.J.
- Cecchi F.,
- Olivotto I.,
- Montereggi A.,
- Santoro G.,
- Dolara A.,
- Maron B.J.
- Kofflard M.J.M.,
- Van Herwerden L.A.,
- Waldstein D.J.,
- et al.
- McKenna W.J.,
- Camm A.J.
- Maron B.J.,
- Olivotto I.,
- Spirito P.,
- et al.
- Elliott P.M.,
- Poloniecki J.,
- Dickie S.,
- et al.
- Dilsizian V.,
- Bonow R.O.,
- Epstein S.E.,
- Fananapazir L.
- Sadoul N.,
- Prasad K.,
- Elliot P.M.,
- Bannerjee S.,
- Frenneaux M.P.,
- McKenna W.J.