American College of Cardiology

MARON AND MCKENNA et al., ACC/ESC Expert Consensus Document on Hypertrophic Cardiomyopathy JACC 2003; 42: 000-000

American College of Cardiology/European Society of Cardiology Clinical Expert Consensus Document on Hypertrophic Cardiomyopathy

A Report of the American College of Cardiology Foundation Task Force on Clinical Expert Consensus Documents and the European Society of Cardiology Committee for Practice Guidelines

The available data (largely from recorded arrhythmic events that triggered appropriate defibrillator interventions) suggest that complex ventricular tachyarrhythmias emanating from an electrically unstable myocardial substrate are the most common mechanism by which SCD occurs in HCM (2,208,209,236,237). Indeed, ventricular arrhythmias are an important clinical feature in adults with HCM. On routine ambulatory (Holter) 24-h ECG monitoring, 90% of adults demonstrate ventricular arrhythmias, which are often frequent or complex, including premature ventricular depolarizations (greater than or equal to 200 in 20% of patients), ventricular couplets (in greater than 40%) or nonsustained bursts of ventricular tachycardia (VT) (in 20% to 30%) (250). Alternatively, it is possible that in some patients supraventricular tachyarrhythmias could trigger ventricular tachyarrhythmias or that bradyarrhythmias occur and require back-up pacing.

It has been suggested that life-threatening tachyarrhythmias could be provoked in HCM by a number of variables either secondary to environmental factors (e.g., intense physical exertion) (23,25) or, alternatively, intrinsic to the disease process. The latter may involve a vicious cycle of increasing myocardial ischemia (190,192–198,238,251) and diastolic (or systolic) dysfunction (37,181–188), possibly impacted by outflow obstruction (13,127), systemic arterial hypotension (29,252,253), or supraventricular tachyarrhythmias (163,244,254) which lead to decreased stroke volume and coronary perfusion.

Although the available data on the stratification of SCD risk are substantial and a large measure of understanding has been achieved, it is important to underscore that precise identification of all individual high-risk patients by clinical risk markers is not completely resolved. This issue remains a challenge due largely to the heterogeneity of HCM disease presentation and expression, its relatively low prevalence in cardiologic practice, and the complexity of potential pathophysiologic mechanisms (1,7,36,41,50,59). Nevertheless, it is possible to identify most high-risk patients by noninvasive clinical markers (21,22,255), and only a small minority of those HCM patients who die suddenly (about 3%) are without any of the currently acknowledged risk markers (21). The highest risk for SCD has been associated with the following (Table 2): 1) prior cardiac arrest or spontaneously occurring and sustained VT (239); 2) family history of a premature HCM-related SCD particularly if sudden, in a close relative, or if multiple in occurrence (5,7,21); 3) identification of a high-risk mutant gene (6,19,63,132,245– 247); 4) unexplained syncope, particularly in young patients or when exertional or recurrent (7,11); 5) nonsustained VT (of 3 beats or more and of at least 120 beats/min) evident on ambulatory (Holter) ECG recordings (240,242,256 –258); 6) abnormal blood pressure response during upright exercise which is attenuated or hypotensive, indicative of hemodynamic instability, and of greater predictive value in patients less than 50 years old or if hypotensive (29,252,253,259); and 7) extreme LVH with maximum wall thickness of 30 mm or more, particularly in adolescents and young adults (21,27,28).

Hypertrophic cardiomyopathy patients (particularly those less than 60 years old) should undergo comprehensive clinical assessments on an annual basis for risk stratification and evolution of symptoms, including careful personal and family history, noninvasive testing with two-dimensional echocardiography (primarily for assessment of magnitude of LVH and outflow obstruction), 24-or 48-h ambulatory (Holter) ECG recording for VT, and blood pressure response during maximal upright exercise (treadmill or bicycle). Subsequent risk analysis should be performed periodically and when there is a perceived change in clinical status.

Recent attention has focused on the magnitude of LVH (as assessed by conventional two-dimensional echocardiography) as an indicator of risk (27). Two independent groups have reported a direct association between magnitude of LV wall thickness and risk of SCD in large HCM populations (21,22,27). In one study (27), extreme LVH (maximum thickness of 30 mm or more), present in approximately 10% of HCM patients, conveyed substantial long-term risk. Sudden cardiac death was most common in asymptomatic or mildly symptomatic adolescents or young adults and was estimated at 20% over 10 years and 40% over 20 years (i.e., annual mortality about 2%). There is supporting circumstantial evidence from retrospective cross-sectional analyses that extreme hypertrophy represents a risk factor for premature SCD because it is observed less commonly in older than in younger patients (21,22,260); this finding could reflect either preferential SCD at a young age, structural remodeling with wall thinning, or both. This relationship of extreme hypertrophy to age is accentuated with wall thicknesses of 35 mm or more, which appear in less than 1% of patients older than 50 years (260). Other investigators, however, have maintained that extreme hypertrophy is a predictor of SCD, only when associated with other risk factors such as unexplained syncope, family history of premature SCDs, nonsustained VT on Holter, or an abnormal blood pressure response during exercise (22). At present, although it is not unequivocably resolved as to whether extreme hypertrophy as a sole risk factor is sufficient to justify a recommendation for prevention of SCD with an ICD, serious consideration for such an intervention should be given to young patients.

The concept that risk of SCD is related to the magnitude of hypertrophy does not, however, infer that the risk is necessarily low when LV wall thickness is less than 30 mm, because other risk markers may be present in a given patient; indeed, the majority of patients who die suddenly do, in fact, have wall thicknesses of less than 30 mm (21,22,27,28). Furthermore, a small number of high-risk pedigrees with troponin T and I mutations have been reported in whom SCD was associated with particularly mild forms of LVH, including a few individuals with normal LV wall thickness and mass (19,248,261). However, such events appear to be uncommon within the overall HCM patient spectrum. Although prognosis is generally not tightly linked to the pattern and distribution of LVH, the preponderance of evidence suggests that segmental wall thickening at the low end of the morphologic spectrum (i.e., less than 20 mm thickness, regardless of its precise location), generally confers a favorable prognosis in the absence of other major risk factors (11,27,28). Such localized hypertrophy includes the nonobstructive form of HCM confined to the most distal portion of LV (“apical HCM”) (33,40,52).

Disorganized cardiac muscle cell arrangement (4,51,236), myocardial replacement scarring as a repair process following cell death (possibly resulting from ischemia due to abnormal microvasculature consisting of intramural small vessel disease or muscle mass-to-coronary flow mismatch) (129,195,199,200,203) and the expanded interstitial (matrix) collagen compartment (262) probably serve as the primary arrhythmogenic substrate predisposing some susceptible patients to re-entrant, life-threatening ventricular tachyarrhythmias. That extreme degrees of LVH can be linked to sudden events is perhaps not unexpected, considering the potential impact of such wall thickening on myocardial architecture, oxygen demand, coronary vascular resistance, and capillary density, all of which thereby create an electrophysiologically unstable substrate. The degree of hypertrophy does not appear to be directly associated with the severity of diastolic dysfunction and limiting symptoms (188,263). Paradoxically, most patients with massive degrees of LVH do not experience marked symptomatic disability (22,27,263), LV outflow obstruction, or left atrial enlargement.

It is a clinical perception that the premonitory symptom most associated with the likelihood of SCD in HCM is impaired consciousness (i.e., syncope or near-syncope) (128,165). However, the sensitivity and specificity of syncope as a predictor of SCD is low, possibly because most such events in this disease are probably not in fact secondary to arrhythmias or related to outflow obstruction. Indeed, there are many potential causes of syncope, some of which are unrelated to the basic disease state and are often neurocardiogenic (i.e., vagal, neurally mediated syndromes) in origin (5,7,11,264). Even when an underlying cause for impaired consciousness cannot be identified, this symptom-complex can be compelling in some HCM patients (128), particularly when it is exertional or recurrent, when it occurs in the young, or in the context of a single recent syncopal episode judged to be disease-related. Therefore, syncope may represent the basis of a defibrillator implant to ensure preservation of life should a life-threatening arrhythmia intervene (208).

Available data suggest that LV outflow obstruction (gradient 30 mm Hg or more) can only be regarded as a minor risk factor for SCD in HCM (29,30,127). The impact of gradient on SCD risk is not sufficiently strong (positive predictive value of only 7%) for obstruction to merit a role as the sole (or predominant) deciding clinical parameter and the primary basis for decisions to intervene prophylactically with an ICD (127).

Identification of HCM in young children is exceedingly uncommon and often creates a specific clinical dilemma because such an initial diagnosis occurring so early in life (frequently fortuitously) raises uncertainty regarding future risk over particularly long time periods. One report suggests that short-tunneled (bridged) intramyocardial segments of left anterior descending coronary artery independently convey increased risk for cardiac arrest, probably mediated by myocardial ischemia (238). However, potential biases in patient selection, the frequency of coronary arterial bridging in surviving adults and those who have died of noncardiac causes, and the need for routine invasive coronary arteriography in order to identify this abnormality prospectively seem to mitigate the potential power of coronary bridging as a risk factor for SCD.

It has been proposed, based on genotype-phenotype correlations, that the genetic defects responsible for HCM could represent the primary determinant and stratifying marker of prognosis and for SCD and heart failure risk, with specific mutations conveying either favorable or adverse prognosis (i.e., high-and low-risk mutations) (6,19,132,138,247,265,266). For example, it has been suggested that some cardiac beta-myosin heavy chain mutations (such as Arg403Gln and Arg719Gln) and some troponin-T mutations are associated with higher incidence of premature death, decreased life expectancy, and early onset disease manifestations, while other HCM genes such as cardiac myosin-binding protein C (particularly InsG791) or alpha-tropomyosin (Asp175Asn) convey a more favorable prognosis (63). However, routine clinical testing for specific mutations believed to be high (or low) risk has been shown to have low yield (265,266). Therefore, it is premature to draw definitive conclusions regarding gene-specific clinical outcomes based solely on the presence of a particular mutation, by virtue of extrapolation from available epidemiologic-genetic data which are formulated from relatively small numbers of genotyped families largely skewed toward high-risk status (6). Consequently, it is becoming increasingly evident that the presence or absence of a particular mutation does not by itself represent sufficient data to convey clear prognostic implications and that HCM mutations may not possess distinctive clinical signatures.

The particular prognosis attached to adult carriers with a mutant HCM gene but without LVH and clinical expression of HCM (54,245), or those individuals who develop hypertrophy de novo in adulthood (6,17,64,65,146), is uncertain; however, at this early juncture, this subgroup would not appear to be associated with an adverse prognosis (Fig. 1). An exception to this tenet may be the small number of SCDs in young people with little or no LVH reported in a very few families with troponin-T mutations (19,245, 247,248).

There is no convincing evidence that invasive markers such as those defined with laboratory electrophysiologic testing (i.e., programmed ventricular stimulation) (264,267) have an important routine role in identifying those HCM patients who have an unstable electrical substrate and are at high-risk for SCD due to life-threatening arrhythmias. Similar to the experience in CAD and dilated cardiomyopathy, polymorphic VT and ventricular fibrillation (VF) (which are the most commonly provoked arrhythmias) are generally regarded as nonspecific electrophysiologic testing responses to multiple ventricular extra-stimuli (5,11), and these specialized laboratory studies are highly dependant on the level of aggression of the protocol (267). For example, stimulation with three ventricular premature depolarizations rarely triggers monomorphic VT in HCM (in contrast to CAD), but frequently induces polymorphic VT or VF, even in some patients at low risk for SCD.

It is now the predominant view that the risk stratification strategies involving laboratory induction of such ventricular arrhythmias are neither desirable in HCM patients on a routine basis nor, per se, justify aggressive intervention (5,7,11). Electrophysiologic studies with or without programmed ventricular stimulation may, however, have some value in selected patients such as those with otherwise unexplained syncope.

Most of the clinical markers of SCD risk in HCM are limited by relatively low positive predictive values due in part to relatively low event rates (11,21,27,28,30,242). However, the high negative predictive values (at least 90%) of these markers suggest that the absence of risk factors and certain other clinical features can be used to develop a profile of patients having a low likelihood of SCD or other adverse events (11). Adult patients can probably be considered low risk if they demonstrate: 1) no or only mild symptoms of chest pain or exertional dyspnea (NYHA functional classes I and II); 2) absence of family history of premature death from HCM; 3) absence of syncope judged to be HCM-related; 4) absence of nonsustained VT during ambulatory (Holter) ECG; 5) outflow tract gradient at rest of less than 30 mm Hg); 6) normal or relatively mild increase in left atrial size (less than 45 mm); 7) normal blood pressure response to upright exercise; and 8) mild LVH (wall thickness less than 20 mm).

Patients with an apparently favorable prognosis in the absence of risk factors constitute an important proportion of the overall HCM population. Most such patients probably will not require aggressive major medical treatment and generally deserve a large measure of reassurance regarding their prognoses. Little or no restriction is necessary with regard to recreational activities and employment, although exclusion from intense competitive sports is advised.

Prevention. Efforts at the prevention of SCD have historically targeted only the minority of patients with HCM in whom SCD risk was unacceptably high. Historically, treatment strategies to prophylactically reduce the risk for SCD or delay progression of congestive symptoms have been predicated on the administration of drugs such as beta-adrenergic blockers, verapamil, and type I-A antiarrhythmic agents (i.e., quinidine, procainamide) to those patients perceived to be at high risk. However, there is no evidence that this practice of prophylactically administering such drugs empirically to asymptomatic HCM patients to mitigate the risk for SCD is efficacious, and this strategy now seems outdated with the current availability of measures that more effectively prevent SCD, such as the ICD. In addition, low dose (less than 300 mg) amiodarone has been associated with improved survival in HCM (243,257), but this agent requires careful monitoring and may not be tolerated due to its potential toxicity over the long risk periods incurred by young patients.

When risk level for SCD is judged by contemporary criteria to be unacceptably high and deserving of intervention, the ICD is the most effective and reliable treatment option available, harboring the potential for absolute protection and altering the natural history of this disease in some patients (208,209,237,268) (Fig. 1). In one multi-center retrospective study, ICDs appropriately sensed and automatically aborted potentially lethal ventricular tachyarrhythmias by restoring sinus rhythm in almost 25% of a high-risk cohort, followed for a relatively brief period of 3 years (208). Appropriate device interventions occurred at a rate of 11% per year for secondary prevention (the implant following cardiac arrest or spontaneous and sustained VT) and at 5% per year for primary prevention (implant based solely on noninvasive risk factors), usually in patients with no or only mild prior symptoms. There was onlya 4to1 excess of ICDs implanted to lives saved. Patients receiving appropriate defibrillation shocks were generally young (mean age 40 years). Implantable cardioverter defibrillators often remained dormant for prolonged periods before discharging (up to 9 years), emphasizing the unpredictable timing of SCD events in this disease, the potentially long risk period, and the requirement for extended follow-up duration to assess survival in HCM studies (26,208). Therefore, while the decision to implant a defibrillator for primary prevention cannot reasonably be deferred beyond the time when high-risk status is first judged to be present, it may precede considerably the time at which the device ultimately discharges. There is an ongoing multicenter international study of HCM patients with ICDs (208) for the purpose of obtaining data on interventional devices in a much larger population over longer periods of time.

The ICD is strongly warranted for secondary prevention of SCD in those patients with prior cardiac arrest or sustained and spontaneously occurring VT (7,208). The presence of multiple clinical risk factors conveys increasing risk for SCD of sufficient magnitude to justify aggressive prophylactic treatment with an ICD for primary prevention of SCD (208,268). Strong consideration should be afforded for a prophylactic ICD in the presence of one risk factor regarded as major in that patient (e.g., a family history of SCD in close relatives) (7,27,268).

Because the positive predictive value of any single risk factor is low, such management decisions must often be based on individual judgment for the particular patient, by taking into account the overall clinical profile including age, the strength of the risk factor identified, the level of risk acceptable to the patient and family, and the potential complications largely related to the lead systems and to inappropriate device discharges. It is also worth noting that physician and patient attitudes toward ICDs (and the access to such devices within the respective health care system) can vary considerably among countries and cultures and thereby have an important impact on clinical decision-making and the threshold for implant in HCM (269). The ACC/AHA/ NASPE 2002 guidelines have designated the ICD for primary prevention of SCD as a class IIb indication and for secondary prevention (after cardiac arrest) as a class I indication (225).

There is, at present, an understandable reluctance on the part of pediatric cardiologists to implant such devices chronically in children (particularly for primary prevention) considering the necessary, ongoing commitment required for maintenance and the likelihood that lead or other (ICD-related) complications will occur over very long time periods. However, while adolescence may represent a psychologically difficult age to be encumbered by an ICD, it should also be emphasized that this is coincidently the period of life consistently showing the greatest predilection for SCD in HCM (7,21–23,25–28,208). One alternative but empiric strategy proposed for some very young high-risk children is the administration of amiodarone as a bridge to later ICD placement after sufficient growth and maturation has occurred. Some investigators also regard the end-stage phase of HCM as a risk factor for SCD, justifying implantation of a cardioverter-defibrillator during the waiting period prior to the availability of a heart for transplant.

Athlete recommendations. In accord with the recommendations of the Expert Consensus Panel of the 26th Bethesda Conference (241), young patients with HCM should be restricted from intense competitive sports to reduce the risk for SCD that may be associated with such extreme lifestyle. A linkage has been established between SCD and intense exertion in trained athletes with underlying cardiovascular disease (including HCM) and SCD (25,270).

There is indirect and circumstantial evidence that the removal of young athletes from the competitive arena reduces risk for SCD (241,271). Not all trained athletes with HCM die suddenly during their competitive phase of life, only some HCM-related SCDs are associated with intense physical activity (25,26), and precision in the stratification of risk for athletes with HCM is particularly difficult given the extreme environmental conditions to which they are often exposed (associated with alterations in blood volume hydration and with electrolytes). Nevertheless, the consensus of the general medical community prudently supports avoiding exposure to most competitive sports for young athletes with HCM to reduce SCD risk, and therefore withdrawal from the athletic arena can be regarded as a treatment modality in this disease (241,271). However, stringent lifestyle or employment modifications for other HCM patients (who are not participants in organized athletics) do not seem justified or practical, although intense physical activity involving burst exertion (e.g., sprinting) or systematic isometric exercise (e.g., heavy lifting) should be discouraged. Although data are scarce, there is presently no evidence to suggest that genetically affected but phenotypically normal family members are generally at increased risk for SCD. Therefore, there is little basis for subjecting such individuals to the same activity restrictions as many other HCM patients, or excluding them from competitive athletics in the absence of cardiac symptoms, family history of SCD, or a mutant gene regarded as malignant. However, periodic (probably annual) noninvasive clinical evaluation directed toward risk assessment is warranted in this subset of patients.