Hypertrophic cardiomyopathy: Sudden cardiac death risk stratification in adults

[first paragraph of article]Sudden cardiac death (SCD) is a devastating and often unpredictable complication of hypertrophic cardiomyopathy (HCM) that may occur as the initial disease presentation, frequently in asymptomatic or mildly symptomatic young people. Until 2000, only small series of patients examining predictors of SCD had been published, with a selection bias towards severe disease. Subsequently, larger series that are more representative of the HCM spectrum have shown that the annual SCD rate is less than 1%, and that there are subgroups of patients with a clearly higher risk.


INTRODUCTION
Sudden cardiac death (SCD) is a devastating and often unpredictable complication of hypertrophic cardiomyopathy (HCM) that may occur as the initial disease presentation, frequently in asymptomatic or mildly symptomatic young people. Until 2000, only small series of patients examining predictors of SCD had been published, with a selection bias towards severe disease. Subsequently, larger series that are more representative of the HCM spectrum have shown that the annual SCD rate is less than 1% 1,2 , and that there are subgroups of patients with a clearly higher risk.

PATHOPHYSIOLOGY OF SCD
Genetic and molecular substrate, myofibrillar disarray, ventricular hypertrophy, microvascular ischemia and fibrosis 3 predispose patients with HCM to re-entrant ventricular arrhythmias 4 . Disruption of intercalated discs, increased myofilament calcium sensitivity and abnormal calcium handling are additional arrhythmogenic factors 5,6 . Life-threatening ventricular tachyarrhythmias can be triggered by a number of environmental factors (e.g., intense physical exertion) or features intrinsic to the disease process, including left ventricular outflow obstruction, systemic arterial hypotension and supraventricular tachyarrhythmias. Systematic analysis of stored electrograms has shown that most ventricular arrhythmias occur spontaneous in normal sinus rhythm, sometimes precipitated by premature ventricular complexes 7 , although rapid atrial fibrillation has also been demonstrated as a trigger 8,9 .

RISK STRATIFICATION AND PREVENTION OF SCD
Implantable cardioverter defibrillators (ICD) are effective in terminating life-threatening ventricular arrhythmias in HCM, whereas pharmacologic therapy has not been demonstrated to provide protection from SCD. ICD implantation in secondary prevention is rarely a clinical challenge 10 , but identifying patients within this heterogeneous disorder who are at high risk of SCD when prior to a first event is a challenge. Clinical guidelines recommend that all patients should undergo SCD risk stratification at their initial evaluation and periodically thereafter. The clinical parameters that associate with SCD are summarized in Table 1 and discussed below.

Prior personal history of aborted ventricular fibrillation or sustained ventricular tachycardia (VT)
Patients with HCM who have experienced sustained ventricular arrhythmia represent the highest risk for subsequent arrhythmic events, with an approximate recurrence rate of 10% per year, although it has been shown that some patients may have no repeated events or have decades-long arrhythmia free intervals between episodes 11,12 .

Family history of SCD
A family history of SCD is associated with an increased risk of SCD in other affected family members, particularly if there are multiple SCD events and if they occurred in young people 13,14 . Nonetheless, discrepancy exists regarding the independent relationship between family history of SCD and risk for the individual patient 15 , probably related to the relative low incidence of events and the variability in the definition of family history of SCD. The average Hazard ratio (HR) from a systematic review was 1.27, 95% confidence interval (CI) 1.16-1.38.

Unexplained syncope
Syncope, defined as a temporary loss of consciousness secondary to transient global cerebral hypoperfusion, is a challenging clinical diagnosis. There are many potential causes of syncope in HCM including sustained ventricular arrhythmias, supraventricular tachycardias, bradyarrhythmias and exercise-related left ventricular outflow tract obstruction (LVOTO). Patients may also experience vasovagal syncope. The difficulty in determining the cause of a temporary loss of consciousness means that syncope alone has a low sensitivity and specificity for SCD. In a systematic review, the average HR of unexplained syncope was 2.68 but the association did not reach statistical significance (95% CI [0.97-4.38]) 16 . Therefore, a careful clinical history is required before it can be considered a potential marker for SCD. It is particularly important when it is exertional or recurrent, it occurs in the young, or in the recent past (<6 months) [17][18][19] .

Maximum left ventricular wall thickness
The magnitude of left ventricular hypertrophy correlates with the risk of SCD 22,23 . This is to be expected given the potential impact of that thickening on myocardial replacement scarring resulting from intramural small vessel disease and mass-to-coronary flow mismatch, creating an electrophysiologically unstable substrate.
Several large studies have shown an independent association between a magnitude of hypertrophy ≥30 mm and SCD 24,25 including a systematic review (HR=3.10, 95% CI [1.81-4.40]) 16 . However, this cut-off value is somehow arbitrary and not based on any specific biological properties and the risk estimate does not abruptly increase for patients with ≥30 mm wall thickness but rather increases in a linear 25 or quadratic fashion 26 and appears to carry more prognostic value in younger patients.

Abnormal blood pressure response during exercise
An inappropriate systemic systolic blood pressure (BP) response during exercise testing, defined as either a failure to increase by at least 20 mmHg or a drop of at least 20 mmHg during effort, is common in patients with HCM 27,28 and seems to be related with LVOTO. It is suggested 29,30 that an inappropriate drop in systemic vascular resistance, despite an appropriate increase in cardiac output is a contributory mechanism. Two studies showed a univariate association between this abnormal response and subsequent SCD 28,31 . However, no study using this definition has demonstrated a significant independent prognostic value. In a systematic review, the association was not statistically significant (HR 1.3, 95% CI [0.64-1.96]) 16 .

Left ventricular outflow tract obstruction (LVOTO)
There is evidence for a higher risk of SCD among patients with left ventricular outflow tract gradients ≥30 mmHg 31,32 and a positive correlation between the magnitude of risk and the severity of obstruction 32 . LVOTO can provoke SCD either by causing severe reduction in cardiac output leading to electromechanical dissociation or by precipitating ventricular arrhythmias though myocardial ischemia caused by increased left ventricular end-diastolic pressure 33 . Conversely, relief of LVOTO through surgical myectomy is associated with low subsequent rates of SCD 34,35 .
Due to its dynamic nature, LVOTO was not originally considered in risk scores. The fact that it can be strongly mitigated by drugs or septal reduction also has the potential to dilute its contribution to SCD.

Age and gender
While there appears to be no difference in SCD rates based on gender, age represents an important factor. SCD is more common in younger patients, especially those under the age of 35 year; however, up to 20% of SCDs occur in patients over the age of 65 36 . As discussed NSVT 30 and left ventricular hypertrophy (LVH) appear more significant as risk factors in younger patients 25 .

Late gadolinium enhancement on cardiac magnetic resonance imaging
Late gadolinium enhancement on cardiac magnetic resonance (LGE-CMR) is used as a surrogate for the degree and distribution of fibrosis, but its positive predictive value for ventricular arrhythmias and SCD in HCM patients is controversial.
LGE-CMR is present in up to 70% of cases, and is associated with impaired systolic function and other risk factors for SCD, in particular NSVT 37,38 . Two recent meta-analysis that assessed prognostic value of LGE-CMR in HCM concluded that it was a predictor of SCD irrespective of LV ejection fraction 39,40 . However, due to its high prevalence, binary analysis of myocardial fibrosis does not aid in decision-making. LGE-CMR extension is also influenced by the imaging protocol used, and therefore, there is no consensus for considering LGE-CMR an independent predictor for SCD.

Left ventricular apical aneurysm
Approximately 2-5% of patients with HCM, typically those with mid-ventricular hypertrophy, develop a thin-walled left ventricular apical aneurysm associated with regional scarring. A higher incidence of clinical events during follow-up have been reported in this subgroup, including a documented risk of SCD of 5% per year 41,42 .

Genetic mutations
Early studies of HCM pedigrees suggested that some mutations in cardiac beta-myosin heavy chain and in troponin-T were associated with a higher incidence of premature death, decreased life expectancy, and early onset disease manifestations than others 43 . However, subsequent studies from unselected consecutive patients did not confirm a clear prognostic association, although most were underpowered to investigate the association between individual mutations and risk. Future advances in genotype-based risk stratification will probably shed light in the management of these patients.

Supraventricular arrhythmias and left atrial size
Atrial fibrillation (AF) and left atrial size reflecting LV filling pressures and remodeling may indirectly reflect disease progression and SCD risk. Left atrial diameter has been associated with SCD in a single study with a HR 1.03; 95% CI [1.00-1.05], p = 0.04 19 . Chronic -but not paroxysmal or persistent -AF has also been associated with SCD 17,31,44,45 .
In a study of 71 HCM individuals with ICD for primary or secondary prevention, ventricular fibrillation or rapid VT episodes (>200 bpm) were more frequently preceded by supraventricular rhythms greater than 100 bpm (p = 0.001) suggesting that supraventricular tachycardia might play a role in the trigger of rapid VT 46 .

Other factors
Electrophysiologic testing with programmed ventricular stimulation has not demonstrated utility in identifying those HCM patients at higher risk for SCD because the induction of VT or ventricular fibrillation is highly dependent on the aggressiveness of the stimulation protocol. Therefore, electrophysiologic testing with programmed ventricular stimulation is not recommended on a routine basis but may be useful in selected patients with unexplained syncope.

GUIDELINES RECOMMENDATIONS FOR ICD IMPLANTATION: 2003 TO 2018
The ACC/ESC Expert Consensus published in 2003 47 recommended ICD in individuals with two or more major risk factors but stated that ''strong consideration should be afforded for a prophylactic ICD'' if a single major risk factor was present. The following major risk factors were considered: prior cardiac arrest or spontaneously occurring sustained VT; family history of a premature HCM-related SCD; unexplained syncope, particularly in young patients or when exertional or recurrent; extreme LV hypertrophy with a maximum wall thickness of 30 mm or more, particularly in adolescents and young adults; abnormal BP response during upright exercise (greater predictive value in patients less than 50 years old or if hypotensive); and NSVT on ambulatory ECG recordings. Minor risk factors were the identification of a high-risk mutant gene, LVOTO, atrial fibrillation, myocardial ischemia and intensive (competitive) physical exertion.
In an update of these guidelines, The ACC/AHA/ESC 2006 guidelines for management of patients with ventricular arrhythmias 48 and the ACC/AHA/HRS 2008 guidelines for device-based therapy of cardiac rhythm abnormalities 49 stated that ICD implantation could be effective or reasonable for patients with HCM who had one or more major risk factors for SCD (both class IIa recommendations, level of evidence C). The 2011 ACCF/AHA guidelines 50 also recommended implantation of an ICD if a single risk factor such as family history of SCD, LV thickness greater than 30 mm or recent unexplained syncope, was present whereas the relevance of NSVT and abnormal BP response was downgraded, and required the concomitant presence of additional risk factors such as left ventricular outflow obstruction, LGE-CMR, apical aneurysm or double mutations to recommend an ICD.
There is agreement that there is a correlation between the sum of risk factors and the incidence of SCD, and that individual risk factors in insolation have poor positive predictive value for SCD. However, validation studies have shown that the simple summation model has a relatively poor predictive performance and leads to the implantation of ICD in patients at low risk 51 . Moreover, previous algorithms have ignored the influence of age and consider variables such as myocardial thickness and LVOTO as binary factors when they are associated with a continuum of risk.
In 2014, the European Society of Cardiology 52 published a new score based on a multicenter, retrospective cohort study that included 3.675 individuals 26 in which risk factors independently associated with SCD in at least one multivariate study analysis were evaluated. Statistical modelling was employed to find clinical variables that were associated with SCD at ≥0. 15  Concerns about the sensitivity of the ESC SCD calculator as compared with the 2011 ACCF/AHA guidelines were raised in a study of 1.629 patients previously risk-stratified according to the ACCF/AHA guidelines. The authors found that the ESC calculator had adequate specificity but poor sensitivity compared to the ACCF/AHA guidelines 55 . However, in a global validation study published in 2018 53 , SCD was observed in 1.4% of individuals in the low risk group (estimated risk of <4%) and in 8.9% of individuals in the high risk group (estimated 5 year risk ≥6%). The number needed to treat (NNT) of ICD required to save one live was 13. The score has shown a markedly higher calibration and discrimination than previous algorithms 54 .
In the latest ESC clinical practice guidelines, LGE-CMR was included as an additional parameter that can be considered in decision-making among patients with intermediate risk score along with the presence of apical aneurysm, LV ejection fraction <50%, double mutations or NSTV during exertion.
The HCM-SCD score is not validated in children or in individuals with myocardial hypertrophy due to metabolic diseases or syndromes with multi-organ involvement. Individuals with LV hypertrophy higher than 35 mm had a low representation in the population of the score, and therefore, the model should be used with caution in this subgroup of patients. The score has not been validated in patients undergoing a septal reduction therapy, and therefore it is not clear whether the prior estimated risk or the recalculated one after therapy should be used. The defined cut-off point for recommending ICD implantation (estimated risk of sudden death >4-6%) is an arbitrary value that has been assumed by balancing medical and socioeconomical criteria.

FINAL CONSIDERATIONS
The decision to recommend ICD implantation is complex. It should be based on individual judgement for the particular patient, by taking into account the overall clinical profile including age, the strength of the risks factors identified, the level of risk acceptable to the patient and family, anxiety, and the potential complications related to the lead systems and to inappropriate device discharges. It is relevant to consider the patient's age, particularly because device complications are more likely in children and young adults over the long period of follow-up 12,56 . Prior to ICD implantation patients should be advised on the risk of inappropriate shocks, implant complications, and the social and occupational implications of an ICD.