Molecular basis of inherited cardiac diseases at risk of sudden death in competitive athletes

Adrenergic stress during competitive sport is a commonly accepted trigger for arrhythmia and SCD, in the presence of underlying inherited cardiac disease such as cardiomyopathy (diseases with structurally and functionally abnormal myocardium in the absence of common condition sufficient to cause it),5 primary arrhythmia syndromes (pure electrical diseases mostly due to ion channel dysfunction)6 or vascular disease (such as Marfan syndrome).7 In an examination of 1866 young athletic SCD's, the most common cause was hypertrophic cardiomyopathy (HCM) (36%), but other inherited diseases were also frequently observed; namely ‘possible’ HCM (17%), arrhythmogenic right ventricular cardiomyopathy (ARVC) (4%), long QT syndrome (LQTS) (4%), aortic rupture (3%) and dilated cardiomyopathy (DCM) (2%).3 A common theme across recent histopathological studies is the under-reporting in the frequency of primary arrhythmia syndromes, as this diagnosis is usually missed by necropsy. For the cardiac geneticists, however, ion channel diseases can be identified via postmortem genetic analyses in up to 10–30% of cases who often demonstrate normal conventional necropsy.8–10

• Download figure
• Open in new tab
• Download powerpoint

Figure 1A

Main phenotypes and genes responsible for cardiovascular inherited diseases with their respective frequency. (A) Cardiomyopathies and (B) rhythm disturbances. This figure is only reproduced in colour in the online version.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1B

continued.

Inherited cardiac disease: genes and inheritance

The main genes involved in inherited cardiac disorders are indicated in figure 1. The mode of inheritance for most of these diseases is autosomal dominant, but in particular phenotypes such as syndromic diseases or paediatric patients with cardiomyopathies, recessive, X-linked or mitochondrial modes of inheritance may also be observed (figure 1A, bottom). The familial aggregation is not always recognised, making it difficult to conclude about a familial/genetic disease. Sporadic cases should be considered as part of genetic disease (usually with autosomal dominant inheritance), especially for HCM and ARVC, either because of a de novo mutation (mutation not transmitted from a parent) or because of incomplete penetrance in the parent (mutation without cardiac expression).11–13

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Strategy for genetic testing in atheletes with a familial context of inherited cardiac disease or with mild abnormalities. ¶Decision by specialized multidisciplinary teams †Case by case approach according to the disease, cardiac features, gene, mutation, specific sport that is practiced. This figure is only reproduced in colour in the online version.

HCM is characterised by the presence of increased ventricular wall thickness or mass in the absence of loading conditions (hypertension, valve disease, etc) sufficient to cause the observed abnormality5 with a disease prevalence of 1/500.14 The usual mode of inheritance is autosomic dominant, characterised by genetic heterogeneity, with more than 900 different mutations described in more than 13 genes15–18 (Sarcomere Protein Gene Mutation Database http://cardiogenomics.med.harvard.edu). These genes encode the proteins that form various components of the contractile apparatus of the sarcomere. The global efficiency of mutation screening in index cases is of 30 to 65%. A study of nearly 200 independent HCM patients reported that 42% of genotype-positive patients had mutations in myosin-binding protein C (MYBPC3), 40% in β-myosin heavy chain (MYH7), 6% in troponin T (TNNT2), 6% in troponin I (TNNI3) and 4% in regulatory myosin light chain (MYL2)(figure 1A).16

ARVC has an estimated prevalence of 1/2000–1/5000 in the general population and is histologically characterised by the replacement of myocytes within the right ventricular myocardium by adipose and fibrous tissue.19 ,20 The disease is transmitted as an autosomal-dominant trait, with five of the responsible genes encoding the major components of the cardiac desmosome (figure 1A). According to the different number of published studies, the global efficiency of mutation detection is 30–70% in index cases. Data from our laboratory of 135 independent French patients indicate that plakophilin-2 (PKP2) is the predominant gene mutation (31% of patients), followed by desmoglein-2 (DSG2, 10%), desmoplakin (DSP, 4.5%), desmocollin-2 (DSC2, 1.5%) and junction plakoglobin (JUP, 0%).21

DCM is defined by left ventricular (LV) dilatation and systolic dysfunction in the absence of abnormal loading conditions or coronary artery disease, sufficient to cause the impairment and prevalence is estimated to be 1/2500.5 ,22 DCM is a sporadic disease in approximately 65% of cases and a familial disease in 35%. Familial forms are characterised by a great genetic heterogeneity (>30 causal genes reported), as well as a huge diversity of related proteins and underlying pathways such as nuclear envelope, cardiac sarcomere, ion channels, transcription factors and dystrophin-associated cytoskeletal complex.12 ,23 ,24 However, the yield of genetic testing for DCM is relatively low, with a global rate of mutation detection of around 20–30% in index cases, and most mutations related to lamin A/C (LMNA, ∼5–10%), β-myosin heavy chain (MYH7, ∼3–7%), cardiac troponin T (TNNT2, ∼4%), myosin-binding protein C (MYBPC3, <1%), sodium channel (SCN5A) and α-myosin heavy chain (MYH6) (figure 1A).25–28

Among the primary arrhythmia syndromes, LQTS is characterised by prolonged QT interval on ECG, syncope and sudden cardiac death due to ventricular tachyarrhythmias.29 ,30 The prevalence is estimated to be 1 in 2500 in the general population31 ,32 but two studies reported a high prevalence of LQTS founder mutations in the Finnish (0.4%) and Norwegian (1%) populations.33 ,34 LQTS is usually of genetic origin, with autosomal-dominant inheritance and an efficiency of mutation identification of 40–60% in index cases.29 ,35 Three common forms of LQTS represent the majority of cases; LQT1, a slow outward delayed rectifier potassium current abnormality encoded by mutations of KCNQ1 gene; LQT2, a rapid component outward delayed rectifier potassium current abnormality encoded by mutation of KCNH2 gene and; LQT3, a gain-of-function sodium channel abnormality encoded by mutation of SCN5A gene. Mutations in KCNQ1 represent 50% of genotyped LQTS, KCNH2 40% and SCN5A accounts for less than 5% of LQTS patients.36 However, it should be noted that some rare mutations have been identified in the auxiliary subunits of ion channels (figure 1B).6 ,37–39

There are several other inherited cardiac diseases that can cause sudden death that the genetic cardiologist may face within the athletic community. Catecholaminergic polymorphic ventricular tachycardia (CPVT) is a rare arrhythmogenic disorder (prevalence 1/10 000) characterised by adrenergic-induced bidirectional and polymorphic ventricular tachycardia.30 ,32 Responsible genes for CPVT encode the cardiac ryanodine receptor (RYR2) and the cardiac calsequestrin 2 (CASQ2; figure 1B). Despite rarely a cause of SCD in athletes, yet worthy of mention in this review are the conditions of Brugada syndrome or short QT syndrome. The global rate of mutation identification in index cases is 65% in CPVT and 20–30% in Brugada syndrome.40

Penetrance of mutations and natural history of the diseases

Penetrance is defined by the proportion of mutation carriers with an obvious clinical expression, such as an abnormal ECG and/or echocardiogram. However, when considering the lifetime cumulative risk of expressing the phenotype, penetrance may be complete (100% risk of developing the disease with age) or incomplete (<100% risk of developing the phenotype). Mutations responsible for inherited cardiac diseases exhibit a variable penetrance, typically increasing with age. The onset of the cardiac expression (phenotype) is therefore usually delayed with a strictly normal cardiac examination for many years, and may explain why some athletes may perform extraordinary physical achievements for many years without any detection of inherited cardiac disease. The general natural history of these diseases is normally characterised by three important progressions; (1) a silent stage without personal symptoms, normal cardiac examinations and a normal ECG/echocardiogram (NB: the mutation is already detectable at this stage), (2) a second stage without symptoms but with detectable cardiac expression of the disease on ECG or echocardiogram (frequently delayed after 20 years of age for the majority of cardiomyopathies) and (3) a third stage with symptoms, obvious cardiac expression of the disease or sometimes the occurrence of unexpected SCD. When considering the lifetime risk of developing the disease, penetrance may be nearly ‘complete’ after 50–60 years of age, in most cases in HCM or DCM (penetrance: ∼95%),11 ,41 whereas for other mutations or diseases such as ARVC, LQT and Brugada syndromes, incomplete penetrance is commonly observed.6 ,42 Furthermore, the age for the phenotypic expression onset is also different according to the disease type, with frequent cardiac expression in young children and even neonates for LQTS but infrequent expression before 10 years of age for HCM or ARVC. Penetrance is also influenced by gender, as the expression of disease is more frequent with earlier onset in males for HCM, DCM, ARVC and Brugada syndrome, whereas LQTS cardiac expression is more frequent in females.6 ,11 ,43

Variable expressivity and genotype–phenotype correlations

Patients with an inherited cardiac disease usually present a variable severity in phenotypic expression, called expressivity, with dramatic differences in the age at onset, degree of symptoms, severity of ECG/echocardiography features, risk of complications and response to medical treatment. This variability is observed in patients sharing the same genetic mutation, both at interfamilial and intrafamilial levels, suggesting that epigenetic factors as well as environmental factors may act on the severity of the phenotype.12

However, the variability in the disease severity is accentuated between families by a great genetic heterogeneity with multiple genes, multiples mutations within the genes, various pathogenic mechanisms of mutations that could be involved in a given patient. The precise underlying gene and mutation may contribute to the variable expressivity of the phenotype,6 ,44 ,45 what is known as phenotype–genotype correlations and may have important consequences for genetic counselling and medical management of the athlete. Such correlations have been observed although they still require to be confirmed by the study of large populations and a prospective follow-up.

In HCM patients, genotype–phenotype correlations have shown that mutations in troponin T (TNNT2) usually cause mild hypertrophy that can escape clinical detection, but are associated with a high risk of premature SCD15 ,46 ,47 whereas mutations in myosin-binding protein C (MYBPC3) have been associated with delayed onset of HCM and significantly lower risk of SCD.48 ,49 In DCM patients, high mortality related to both heart failure and premature SCD have been described in patients with LMNA mutations.50 In ARVC patients, a high risk of left ventricle dysfunction was associated with desmoplakin (DSP) or desmoglein-2 (DSG2) mutations.21 ,51 In LQTS patients, cardiac events are often determined by the underlying genotype. In LQT1 phenotype, involving mutations in KCNQ1, cardiac events are mostly triggered by adrenergic stimulation, typically exercise. In LQT2 phenotype, involving mutations of KCNH2, cardiac events are triggered by emotion and in LQT3 patients with mutations in SCN5A cardiac events are triggered by sleep/rest. Efficacy of β-blocker treatment is also variable in LQTS patients according to the underlying gene: very efficient to reduce syncope and SCD in LQT1, whereas less efficient in patients with SCN5A mutations.6 ,52–54

Not only the gene but also the precise mutation within a given gene might be important. For example, in LQT1 patients, the higher risk of SCD concerns patients with mutations in the transmembrane domains of the protein encoding for the pore of the channel55 ,56 and in DCM patients, a higher risk of SCD was associated with the LMNA mutations that led to truncated proteins (mutations disrupting the reading frame during translation).50

Recent molecular findings also shed light on the intrafamilial variability of these diseases, especially in HCM and ARVC. Even in the context of familial forms and a dominant inheritance, a subset of the patients (3–10%) may carry more than one heterozygous mutation, so-called multiple mutations (either in the same gene or in another gene). Interestingly, these patients exhibit a more severe phenotype and a more frequent risk of SCD.21 ,57–59 Additional factors may also modulate the phenotype and contribute to the intrafamilial variability, such as common genetic polymorphisms.60 ,61 or acquired/environmental factors (such as myocarditis), but they are still poorly understood and further research is required in this specialised area.

Genetic testing in practice

Genetic testing was developed in some authorised medical laboratories for the purpose of routine use in order to improve both genetic counselling and the medical management of patients and relatives with inherited cardiac pathologies,11 ,26 ,62 ,63 and these basic principles are applicable to all populations including high-level athletes.

The molecular testing begins with a blood sampling of an index patient with an obvious form of a disease (with a certain or possible genetic origin), for which genes of interest are tested for mutation by sequencing (determination of the primary sequence of genes). For most cardiac diseases, the analysis concerns several genes and all the coding sequences of each of these genes. The process can therefore be quite long-winded and results for the index patient may often take several months. In HCM and ARVC, genes are tested according to their frequency but in patients with a family history of SCD, TNNT2 can be first tested. A suggestive phenotype can orientate the screening, as in DCM with conduction defect where the LMNA gene is the first tested and in LQTS patients where the ECG guides the testing. The usual efficiency of a conventional mutation screening in a medical laboratory may therefore vary from quite high (50–70% in HCM, LQTS or CPVT and ∼50% in ARVC) to quite low (20–30% in DCM or Brugada syndrome). The efficiency of the mutation screening depends not only on the complexity of the underlying molecular data, that is, the number of genes to screen, the large size of these genes and the sensitivity of the methods, but also on a careful characterisation of the phenotype and the possibility of phenocopy or environmentally induced phenotype mimicking a genetically determined trait. For example, a patient with concentric HCM associated with a Wolf Parkinson White syndrome will be tested first on PRKAG2 gene before sarcomeric genes.

In a few cardiac diseases with a particular phenotype, the analysis can be focused on one or few more predominant mutations within a given gene as in haemochromatosis due to mutations in the hereditary haemochromatosis protein encoded by the HFE gene, familial amyloidosis with mutations in transthyretin or Marfan syndrome with mutations in fibrillin-1 gene. The result is then available in few days or weeks with a very high efficiency.

Conversely, when the causal mutation is identified in the index-patient of the family, it is easy (one PCR/sequence reaction), quick (within few days) and efficient (100% result: the familial mutation is present or absent) to determine the genetic status of any relative within this family.

The recent evolution of sequencing technologies and their use in the very near future for routine diagnosis will dramatically change the practice and interpretation of genetic testing. Next Generation Sequencing (NGS) is used to designate new methods of sequencing that provide massive high throughput sequencing in a short time.64 ,65 NGS can be used for the simultaneous analysis of numerous target disease or candidate genes for a given disease. In inherited cardiomyopathies and rhythm disturbances, 50–70 disease genes could be analysed simultaneously. This approach will undoubtedly revolutionise the diagnostic approach of genetic medicine but it also poses many new challenges, not least the analysis of massive amounts of data and the interpretation of the many tens or thousands genetic variants that will be identified in individual patients or athletes.

Genetic testing and the interpretation of results in athletes

In the case of an athlete with a suspected cardiomyopathy or ion channelopathy, there are two different situations for the recommendation of mutation analysis to be performed (figure 2).

The first situation is when there is a family history of cardiac hereditary disease and the mutation is identified in the athlete's family. Genetic testing can be proposed to the athlete (asymptomatic with a normal or borderline cardiac examination) in order to search for the presence or absence of the specific mutation previously identified in their family. The process leads to a rapid and appropriate answer whereby (1) the presence of the mutation in the athlete is identified, leading to complete cardiac examination, and the discussion on specific preventive therapy or life-style modifications, together with effective management and regular medical follow-up or (2) the identified mutation in the family is absent within the athlete. Thus, the athlete can be discharged without the need for further cardiac examination; only the basic risk of the disease prevalence in the general population remains. The second situation is when there is no family history of cardiac hereditary disease, but the athlete may present borderline cardiac abnormalities on ECG or echocardiography and/or symptoms, however not sufficient per se to confirm a diagnosis of a cardiomyopathy or primary cardiac arrhythmia. In such cases, where there is adequate evidence to consider the hypothesis of an inherited disease, genetic testing can be proposed directly to the athlete.

If a mutation is identified, then the diagnostic information is crucial for the athlete and their family, leading to appropriate cardiac examination and long-term management. The interpretation of the genetic variant is, however, a very important step that requires several criteria to be studied66 including (1) the absence from an ethnically matched control population of at least 400 chromosomes and from web access polymorphisms data bases; (2) the nature of the mutation, where non-sense mutations as insertions, deletions and stop codons are usually pathogenic verses substitution of amino acids that may be questionable and (3) in silico modelisation of the consequences (location of the mutated residue in a conserved amino acid sequence, predicted secondary structure regions from published reports and the databank such as UniProtKB/Swiss-Prot). If no mutation is identified in this second situation, no conclusive diagnosis can be drawn since a mutation is not always identified in patients with obvious familial disease. Finally, if only a genetic variant of unknown significance is identified in this second situation, then no conclusion can be drawn. The interpretation of the result in this second scenario is therefore very important and it should be borne in mind that only a positive result (identification of a pathogenic mutation) is meaningful.

Risk of cardiac death at the early stage of inherited cardiac diseases

The risk of cardiac complications in patients with an obvious form of inherited cardiac disease is well known, as well as the deleterious role of intensive or competitive sport practice through the occurrence of ventricular arrhythmia. For athletes, specific therapeutic management often includes high-intensity sporting restriction, as underlined by international consensus recommendations.67–70 In contrast, however, the risk of a cardiac event at an earlier stage, in athletes carrying a causal mutation but without obvious cardiac expression of the disease, is poorly understood.

Management of athletes with a mutation but a normal or borderline cardiac examination

To evaluate the risk of SCD associated with competitive sports in the presence of a cardiomyopathy or primary arrhythmia syndrome is a dilemma for the attending cardiologist. This has led to several international recommendations for the preparticipation screening in competitive sports (focused on the identification of a cardiac disease not recognised until now),84 ,85 as well as recommendations for participations of patients (with a recognised cardiac disease) for competitive sports or leisure time physical activities.67 ,68 ,86–89 The practical management of athletes however remains a challenging issue, as little longitudinal data exist supporting the beneficial role of sporting disqualification upon death rates of athletes identified with an inherited cardiac disease, outside of the Italian experience. Because of this death of data, sporting disqualification should also be viewed as the last intervention available in the cardiologist's armoury. It can be anticipated that with future improvement in the understanding of such diseases and their potential risk of adverse cardiac events associated with intensive sport, the objective should and will be to reduce the number of unnecessary disqualifications and to modulate rather than stop the sports activity of the athlete in question.

It is important to realise that the decision for sporting disqualification is based upon the stage of the disease when it is identified within the athlete. This is due to the fact that the natural history of inherited cardiac diseases is often characterised by a dynamic process and a continuum from (1) an asymptomatic mutation carrying athlete in the absence of cardiac expression or who demonstrates mild-to-borderline abnormalities, (2) an asymptomatic athlete with obvious cardiac expression with or without specific SCD risk factors such as non-sustained ventricular tachycardia in HCM or extremely prolonged QTc interval in LQTS patient and finally (3) an athlete with obvious cardiac expression who presents the occurrence of symptoms and/or cardiac complications. Although the risk of SCD is not always strictly related to the stage of cardiac expression, there is a general consensus on the disqualification of competitive sport in athletes with obvious cardiac expression.68 ,86 ,87 ,89 These recommendations are supported by the reduced number of sudden deaths in Italy since the development of systematic preparticipation screening in athletes.90

The management of asymptomatic athletes who carry a mutation but without any cardiac expression or with only borderline abnormalities, is a particularly challenging issue without consensus.68 The risk of sudden death at this early stage is probably low but not negligible, although available data are limited. In the context of such ‘genotype-positive–phenotype-negative’ athletes, there are significant differences between European and North American policy. Study Groups of the European Society of Cardiology recommend86 the disqualification of all mutation carriers (especially within HCM or Marfan athletes), whereas the North Americans (Task Force 4) authorise full competitive sport (except for swimming in individuals with LQT 1 mutation).86 ,87

In our institutions, the management of genotype-positive–phenotype-negative individuals and the subsequent discussion regarding sporting restriction or clearance is based on a case-by-case approach taking into account several factors: (1) the type of gene and mutation identified in the athlete, as some mutations have been acknowledged as being at either a high or low risk for an adverse cardiac events; (2) the familial history and number of premature cardiac deaths as several sudden deaths before 50 years of age in a family represent a known risk factor for HCM and can be considered important in the discussion of sporting restriction with the athlete. Unfortunately, for most primary arrhythmia syndromes, these scientific evidence is lacking; (3) the type of sporting activity undertaken, as highly dynamic sports have been linked with SCD particularly soccer and basketball for HCM and swimming for LQT and finally (4) the presence in the athlete of personal SCD risk factors after prognostic stratification such as non-sustained ventricular tachycardia in HCM mutation carriers.

The following two cases examples from our institutions illustrate the clinical value of genetic testing in the diagnosis of the condition, thus helping in the discussion of sporting restriction or medical clearance for the athlete. One case was brought in as an athlete with symptoms, whereas the second case is ‘predictive’ in the context of a recognised inherited disease in the athlete's family.