1 Genetic Testing and Cardiac Channelopathies

Genetic diagnostics were proposed 20 years ago when the first associated channelopathy genes were discovered (Wilde and Behr 2013). Today, genetic information is used to confirm a clinical diagnosis and as a powerful preventive tool to identify family members at risk of developing electrical disorders despite having a normal ECG. Genetic diagnostics are currently performed on six highly phenotypically characterized channelopathies such as Brugada syndrome (BrS), cardiac conduction defects (CCD), long QT syndrome (LQTS), short QT syndrome (SQTS), catecholaminergic polymorphic ventricular tachycardia (CPVT) and arrhythmogenic right ventricular cardiomyopathy (ARVC). These channelopathies present specific ECG indices at baseline or are unmasked by provocative tests that are often coupled with a singular circumstance of an arrhythmic event outcome. Such clinical information guides genetic testing towards candidate genes harbouring the highest rate of causal mutations. Therefore, the added value of molecular diagnostics could be inherent in the preventive screening of relatives who are either negative or borderline after clinical evaluation. Furthermore, according to the gene identified as being associated with the channelopathy, therapy and care can be adapted accordingly (Imbrici et al. 2016). However, in certain cases, such as young patients presenting idiopathic ventricular fibrillation or (resuscitated) sudden cardiac death, neither ECG indices nor relatives can be identified. In this context, genetic testing could uncover variant(s)/gene(s) and guide clinicians to a diagnosis. However, extreme caution must be taken vis-à-vis our ability to interpret rare genetic variants without supporting clinical and functional or familial segregation data (Wijeyeratne and Behr 2017).

2 Present Context of Genetic Testing

In the last few years, the number of genes associated with arrhythmias has considerably increased. However, the major genes that account for a large fraction of the cases remain limited. “Minor” genes are usually poorly characterized in terms of function and pathophysiological role. As a consequence, the identification of variations in these genes often leads to results that are difficult to interpret. Genetic testing interpretation is also particularly challenging owing to the overlapping, variable and incompletely penetrant nature of the clinical presentations of the channelopathies.

Next-generation sequencing (NGS) technologies rapidly developed in diagnostic laboratories and have become a routine approach. However, the sequencing scale is a matter of debate. Indeed, the technologies allow to focus the screening on panels of 20–200 genes, explore the whole coding region of the genome (whole exome sequencing (WES)) or even give access to the whole genome of a patient. Target resequencing panels (TRP) is commonly employed for diagnostic purposes given the reasonable cost, high sequencing coverage and limited interpretation and ethical issues. Different TRP strategies can be applied from a cardiac panel with 60–180 genes covering “all” (channelopathies and cardiomyopathies) known genes or an arrhythmia panel focused on 20–60 genes directly associated with genetic electrical defects or even a restricted panel containing only the “key” genes corresponding to the most prevalent cardiac arrhythmia genes. The use of small panel focusing on well-characterized genes aims to reduce the number of genes tested and then limit the identification of variant with uncertain interpretation. The pros and cons related to the different approaches using NGS technologies are discussed further.

3 Genetic Testing for Inheritable Cardiac Channelopathies: Daily Practice

3.1 Brugada Syndrome (BrS)

3.1.1 Clinical Description

Brugada syndrome (BrS) was first described in 1992 (Brugada and Brugada 1992) and was based on a familial form of ventricular fibrillation associated with a singular ECG pattern. The latest guidelines define a diagnosis of BrS diagnostic as follows: “patients with ST-segment elevation with type 1 morphology ≥2 mm in one or more leads among the right precordial leads V1 and/or V2 positioned in the second, third, or fourth intercostal space, occurring either spontaneously or after provocative drug test with intravenous administration of sodium channel blockers” (Priori et al. 2015). The prevalence of BrS is estimated at approximately 0.05% in Europe and is higher in Asia (0.12%)—particularly in Thailand, the Philippines and Japan. The mean age on diagnosis is 40. BrS predominantly affects (80%) males after puberty (Andorin et al. 2016) and is associated with sudden cardiac death occurring mostly at rest (Tomaselli and Barth 2016).

3.1.2 Genetic Testing

The first gene identified was the SCN5A gene encoding for the major cardiac sodium channel NaV1.5 (Chen et al. 1998). The prevalence of mutations in SCN5A is from 20 to 25% of the cases and remains the major gene in BrS (Crotti et al. 2012; Le Scouarnec et al. 2015). So far, 24 genes have been described as harbouring rare genetic variations in BrS patients (Table 13.1). They can be categorized in three main groups according to whether they affect the sodium (I Na; SCN5A, GPD1L, SCN1B, SCN3B, RANGRF, SCN2B, PKP2, SLMAP, SCN10A and FGF12), potassium (I K; KCNJ8, KCNH2, KCNE3, KCND3, KCNE5, KCND2, SEMA3A, ABCC9 and KCNAB2) or l-type calcium (I Ca-L; CACNA1C, CACNB2B and CACNA2D1) currents. The last two associated genes affect non-selective channels (HCN4 and TRPM4). It should be noted that among the BrS-associated genes, only two (GPD1L and KCNAB2) have been uncovered by the powerful and hypothesis-free familial approach. Moreover, apart from the SCN5A gene, the contribution of other genes remains extremely low or uncertain for some of them (Fig. 13.1) (Le Scouarnec et al. 2015).

Table 13.1 Genes identified in Brugada syndrome (BrS)
Full size table
Fig. 13.1
figure 1

Decision tree model for genetic testing in Brugada syndrome

Full size image

The autosomal dominant model was the first genetic model proposed. However, the segregation study of SCN5A mutations among BrS pedigrees revealed a low penetrance and a variable expressivity of the pathology suggesting modulating factors (Probst et al. 2009). Despite the low predictive value of SCN5A status, the genotyping of this sodium channel remains the gold standard for BrS genetic testing since loss of function of the NaV1.5 channel has been clearly implicated in BrS pathophysiology (Tan et al. 2003). Furthermore, clear evidence exists describing the genotype/phenotype relationship between SCN5A carriers and conduction defects increasing the risk of arrhythmia (Probst et al. 2010). All in all, approximately 300 distinct SCN5A mutations have been described in patients with the disorder. Of these, two-thirds are missense variations and one-third are nonsense mutations, splice-site mutations and small insertions/deletions that lead to a truncated channel protein (Kapplinger et al. 2010). Other genes can be screened secondarily if SCN5A screening remains negative (see decision tree in Fig. 13.1), but extreme caution must be taken for variant interpretation since little evidence has been noted until now (Le Scouarnec et al. 2015).

Genetic testing plays a major role in the presymptomatic screening of BrS relatives. In the context of BrS families presenting a SCN5A mutation, genetic screening turns out to be of interest especially when ECG (even after provocative test) fails to diagnose the patient (Probst et al. 2009). Presymptomatic genetic testing appears to be even more pertinent in the context of early childhood when provocative tests using sodium blockers are not systematically performed for questionable relevance. Moreover, episodes of fever are frequent in the young, increase the risk of arrhythmia and emphasize the importance of identifying such individuals. Of note, a recent study on young BrS cases shows that genetic testing can uncover a higher proportion of SCN5A mutations (47%) than in adults—supporting the important role of genetic testing.

About 70% of BrS cases remain negative after genetic testing. The first explanation could be the fact that variants within the candidate genes cannot be detected with the sequencing technologies. Indeed, large rearrangements such as large deletion or insertion [also called copy number variation (CNV)] require additional methods for detection such as MLPA (multiplex ligation-dependent probe amplification). This investigation appears to be more and more performed (Selga et al. 2015; García-Molina et al. 2013; Koopmann et al. 2007), but few CNV have been reported so far (Eastaugh et al. 2011). Another hypothesis would be that additional genes could remain uncovered. Furthermore, a recent genome-wide association study suggested that BrS could follow a more complex genetic model than the Mendelian model commonly applied with the combination of common and rare variants of different size effects influencing the risk of developing BrS (Bezzina et al. 2013). Another hypothesis to explain the two-thirds of BrS cases with missing molecular diagnosis could be the presence of acquired Type I BrS ECG in the general population. Drugs have been identified as capable of inducing Type I BrS ECG—especially psychotropic and analgesic-anaesthetic drugs (Konigstein et al. 2016). An up-to-date list of drugs is accessible online to warn BrS patients about increasing their risk of arrhythmia with such drugs (http://www.brugadadrugs.org/) (Postema et al. 2009).

3.2 Long QT Syndrome (LQTS)

3.2.1 Clinical Description

Congenital long QT syndrome (LQTS) is a group of cardiac “channelopathies” characterized by delayed ventricular repolarization manifesting as QT interval prolongation on the ECG in the setting of a structurally normal heart (Morita et al. 2008). The prevalence of LQTS varies from 1/2000 to 1/5000 (Goldenberg and Moss 2008; Schwartz et al. 2009), with a female predominance (2/3 of the patients) (Imboden et al. 2006). LQTS also shows variable expressivity and incomplete penetrance (Roden 2008). The first descriptions of the disease were provided by Jervell and Lange-Nielsen in 1957 (Jervell and Lange-Nielsen 1957) and by Romano and Ward (Ward 1964; Romano et al. 1963). Electrocardiographic findings are characterized by the presence of a prolonged heart rate-corrected QT interval on ECG (QTc). LQTS can be diagnosed in the presence of a QTc > 480 ms (Priori et al. 2015). The length of the QT interval is associated with the risk of syncope and sudden death. There is a high risk when QTc > 500 ms and an extremely high risk when QTc > 600 ms (Goldenberg and Moss 2008). The presence of T-wave alternans despite proper therapy is a sign of electrical instability and requires preventive measures. Patients with syncope or cardiac arrest before the age of 7, especially in their first year of life, have a higher risk of arrhythmias and sudden death (Priori et al. 2004; Spazzolini et al. 2009).

3.2.2 Genetic Testing

Genetic testing is part of the diagnostic criteria of LQTS (Priori et al. 2013). LQTS is most often inherited in an autosomal dominant manner (Romano-Ward syndrome) (Schwartz et al. 1993) and rarely in an autosomal recessive manner associated with sensorineural deafness (Jervell and Lange-Nielsen syndrome).

To date, more than 1200 pathogenic variations have been identified in 17 different genes (KCNQ1, KCNH2, SCN5A, ANK2, KCNE1, KCNE2, KCNJ2, CACNA1C, CAV3, SCN4B, AKAP9, SNTA1, KCNJ5, CALM1, CALM2, TRDN and TECRL) (Table 13.2) (Tester and Ackerman 2014). A causal mutation is found in these genes in 80–85% of LQTS patients. However, LQTS1 (KCNQ1; 35% of the cases), LQTS2 (KCNH2; 30% of the cases) and LQTS3 (SCN5A; 10% of the cases) comprise 90% of the LQTS mutations (Fig. 13.2) (Tester and Ackerman 2014, 2008). The majority of LQTS-causing mutations are coding-region single-nucleotide substitutions or small insertions/deletions. However, a few large gene rearrangements involving single or multiple exons deletions/duplications have been described (Barc et al. 2011).

Table 13.2 Genes identified in long QT syndrome (LQTS)
Full size table
Fig. 13.2
figure 2

Decision tree model for genetic testing in long QT syndrome

Full size image

LQTS is probably the heritable arrhythmia syndrome for which the genotype-phenotype relationship has been the most understood in terms of clinical manifestations, risk stratification and response to therapy. The phenotype-genotype association could facilitate phenotype-directed genetic testing. Indeed, genotype can be inferred by thorough clinical evaluation: swimming, physical exertion or emotional stress cardiac events strongly indicate mutations in KCNQ1. Cardiac events triggered by auditoria stimulation strongly indicate mutations in KCNH2, and symptoms at rest or during sleep are generally observed in LQT3 (Schwartz et al. 2001). T-wave morphology could also help to distinguish different LQTS subtypes: LQT1 exhibits a broad-based T-wave, LQT2 exhibits a low-amplitude notched or biphasic T-wave and LQT3 exhibits a late-appearing T-wave. These gene-suggestive ECG patterns could be helpful in guiding genetic testing, but exceptions to these relatively gene-specific T-wave patterns exist. Of note, from 20% to 25% of the patients with a long QT syndrome confirmed by the presence of a mutation could have a normal QTc (Priori et al. 2003; Goldenberg et al. 2011).

Moreover, the risk of arrhythmia is higher in LQT1 and LQT2 patients with a QTc > 500 ms and LQT3 in men whatever the interval QT duration. Mutation location can also be interpreted: mutations in the cytoplasmic loops of LQT1, mutations with dominant negative ion current effects and mutations in the pore-loop region of LQT2 have been associated with a different risk of cardiac events (Barsheshet et al. 2012; Moss et al. 2002; Shimizu et al. 2009; Migdalovich et al. 2011). In contrast, mutations in the C-terminal region tend to be associated with a milder phenotype (Donger et al. 1997; Crotti et al. 2007).

Genetic testing for the major forms of LQTS (KCNQ1, KCNH2 and SCN5A) has a major role in the diagnosis of probands, risk stratification, familial screening and treatment and could be proposed soon after birth. This is applicable only for pathogenic variations. However, many variations are classified as variants of unknown significance (VUS) and remain uninterpretable and therefore have no clinical impact (Steffensen et al. 2015). In contrast, 15–20% of patients have negative genetic tests, but that does not rule out the presence of the disease. Mutations in other genes (LQT9 to 13) have been identified in a few patients, and their prevalence in LQTS requires evaluation (Schwartz et al. 2012).

Very recently, mutations in calmodulin (CALM1, CALM2), triadin (TRDN) and TECRL genes have been associated with highly lethal arrhythmias in the settings of LQTS and catecholaminergic polymorphic ventricular tachycardia (CPVT) (Altmann et al. 2015; Pipilas et al. 2016; Devalla et al. 2016; Nyegaard et al. 2012; Crotti et al. 2013; Marsman et al. 2014).

3.2.2.1 Syndromic Forms of LQTS

Syndromic forms of LQTS have also been reported as presenting a broader cardiac phenotype than QT interval prolongation or extracardiac abnormalities.

3.2.2.1.1 LQT4: Ankyrin-B Syndrome

Ankyrin-B syndrome presents a complex cardiac phenotype including QT interval prolongation, bradycardia, sinus node dysfunction, atrial fibrillation and ventricular arrhythmia (Schott et al. 1995; Le et al. 2008). Of note, the ANK2 gene encodes for a non-channel protein but is essential for targeting and stabilization of structural proteins and ion channels. Remarkably, most of the mutations are gathered within the death/C-terminal domain of the protein (Hashemi et al. 2009).

3.2.2.1.2 Lange-Nielsen Syndrome

A recessive form of LQTS, the Jervell and Lange-Nielsen syndrome is caused by homozygous or compound heterozygous mutations in KCNQ1 or its auxiliary subunit KCNE1. It is characterized by a congenital deafness, a prolonged QTc on ECG (>500 ms) and ventricular tachyarrhythmias. This syndrome is very rare (1/200,000 to 1/1,000,000) but virulent, and in 50% of the cases, the pathology manifests before the age of 3 (Pagon et al. 1993).

3.2.2.1.3 Andersen-Tawil Syndrome (ATS)

The Andersen-Tawil syndrome is characterized by periodic paralysis, a prominent U wave, a long QT interval with ventricular arrhythmias predisposing to sudden cardiac death and dysmorphic features: short stature, scoliosis, low-set ears, widely spaced eyes, small mandible, clinodactyly, brachydactyly and syndactyly. ATS is inherited in an autosomal dominant manner with variable penetrance. Mutations in the KCNJ2 gene, encoding the alpha subunit of the Kir2.1 potassium channel, are implicated in 60% of the cases. A mutation in the KCNJ5 gene, encoding the alpha subunit of the Kir3.4 potassium channel, has also been described (Kokunai et al. 2014).

3.2.2.1.4 Timothy Syndrome (TS)

The Timothy syndrome is caused by mutations in the CACNA1C gene (LQT8) encoding the alpha subunit of the CaV1.2 calcium channel. The syndrome often results from de novo mutation or from germinal mosaicism in one of the parents. Genetic testing of this syndrome is singular owing to its genetic homogeneity since the cases described so far have been carrying the p.G406R or p.G402R gain-of-function mutations (Tester and Ackerman 2014). This is a multisystem disease characterized by cardiac, hand/foot, facial and neurodevelopmental features. Typical cardiac findings include a rate-corrected QT interval >480 ms, functional 2:1 AV block with bradycardia, tachyarrhythmias and congenital heart defects. The diagnosis of Timothy syndrome is generally made within the first few days of life although it may be suspected prenatally by identifying 2:1 AV block or bradycardia in the foetus (Pagon et al. 1993).

Thus, based on the LQT clinical presentation and potential associated cardiac or not cardiac symptoms, genetic testing is proposed to be performed according to the decision tree presented in Fig. 13.2.

3.3 Short QT Syndrome (SQTS)

3.3.1 Clinical Description

SQTS was first described in 2000 and remains one of the rarest channelopathies (Gussak et al. 2000). Prevalence is difficult to estimate because of the limited number of patients (Gollob et al. 2011). SQTS leads to palpitation, syncope and sudden cardiac death, typically during childhood. It is characterized by a short QT interval on the ECG with peaked T-wave and a high risk of fatal arrhythmias (Patel et al. 2010).

A complete consensus of the cut-off value of QT interval does not exist yet. Diagnostic criteria have been proposed: for QTc <330 ms or <360 ms (350 ms in men and 365 ms in women) if associated with a pathogenic mutation, a family history of SQTS, VT/VF with normal structural heart or of sudden death before the age of 40 (Gollob et al. 2011; Bjerregaard 2011; Veltmann and Borggrefe 2011). Of note, patients with SQTS present a fairly similar QT interval whatever the heart rate variations requiring no or very little heart rate correction. Patients with the shortest QT duration present the highest risk of rhythmic events (Giustetto et al. 2011).

An implantable defibrillator is indicated in high-risk SQTS patients who have experienced VT/VF or resuscitated sudden death and in patients with a familial story of cardiac sudden death. Treatment with quinidine could be an alternative because of its prolongation of QT duration (Priori et al. 2015; Giustetto et al. 2011; Gaita et al. 2004).

3.3.2 Genetic Testing

Pathogenic variations have been described in 6 genes (KCNQ1, KCNH2, KCNJ2, CACNA1C, CACNB2B and CACNA2D1) harbouring mutations that mirror the functional effect of those encountered in LQTS (Brugada et al. 2004; Bellocq et al. 2004; Priori et al. 2005; Antzelevitch et al. 2007). Mutations in potassium channel genes produce a gain of function, and in contrast, mutations in the calcium genes produce a loss of function resulting in shorter AP durations. Mutations in these genes account for 50% of SQTS cases according to the literature, but in clinical practice, the diagnostic yield of genetic testing is very modest in these genes, which supposes that there are other causal genes and that a negative genetic cannot exclude the diagnosis.

3.4 Catecholaminergic Polymorphic Ventricular Tachycardia (CPVT)

3.4.1 Clinical Description

Catecholaminergic polymorphic ventricular tachycardia (CPVT) is a hereditary arrhythmogenic disease that is responsible for familial forms of sudden death in infants and young adults. The prevalence for CPVT in the general population is estimated to be 1/10,000 (Van der Werf and Wilde 2013). Arrhythmias occur as recurrent syncope triggered by adrenergic stimulation such as emotional or exercise stress, while baseline ECG is often normal and cardiac imaging reveals no structural abnormality (George et al. 2007). Approximately, 30% of patients with CPVT develop symptoms before 10 years of age and 60% before 40 years (Liu et al. 2008; Ackerman et al. 2011). A family history of exercise-related syncope, seizure or sudden death is reported in 30% of the patients. As with LQTS1, swimming is a potentially lethal arrhythmia precipitating CPVT, and the symptoms are often attributed to alternative diagnoses, most commonly epilepsy. Drug challenge with epinephrine or isoproterenol may elicit arrhythmias. In the absence of beta-blocker treatment, VT/VF and syncope are associated with a high rate of recurrence with 30–50% mortality before the age of 40 years (Hayashi et al. 2009). Beta blockers, combined with restriction of exercise and avoidance of stressful situations, are the first-line therapy for CPVT. All patients and family members who harbour a disease-causing mutation, irrespective of symptoms, should be treated with beta blockers. Left cardiac sympathetic denervation or ICD implantation could be effective to prevent recurrent VT/VF and SCD in patients intolerant to beta blockers or for those with persistent symptoms or arrhythmias despite treatment (Wilde et al. 2008). However, CPVT patients appear to be very sensitive to catecholamines, and the subsequent ICD shock has a high probability of initiating a vicious cycle with multiple shocks caused by recurrent ventricular tachycardia or VF (Wilde et al. 2008).

3.4.2 Genetic Testing

The presence of a pathogenic mutation is of diagnostic value. Two genetic models of CPVT have been identified: an autosomal dominant form due to mutations in the cardiac ryanodine receptor 2 (RYR2) gene and less common autosomal recessive forms due to mutations in the CASQ2, TRDN and more recently TECRL genes (Table 13.3) (Devalla et al. 2016). Mutations in RYR2 are found in 50%–60% of CPVT cases, whereas mutations in CASQ2 and TRDN are found in less than 10% of CPVT patients (Fig. 13.3) (Roux-Buisson et al. 2012; Rooryck et al. 2015). Rarely, mutations are found in calmodulin genes (CALM1, CALM2 and CALM3), the KCNJ2 gene (Andersen-Tawil syndrome also known as LQT7) and the ANK2 gene (also known as LQT4) (Nyegaard et al. 2012; Mohler et al. 2004; Tully et al. 2015). A locus associated with an early-onset lethal form of recessive CPVT was found on chromosome 7p14-p22, but so far no gene has been identified (Bhuiyan et al. 2007).

Table 13.3 Genes identified in catecholaminergic polymorphic ventricular tachycardia (CPVT)
Full size table
Fig. 13.3
figure 3

Decision tree model for genetic testing in catecholaminergic polymorphic ventricular tachycardia

Full size image

Interestingly, RYR2 mutations cluster in both N- and C-terminal domains of the protein and within transmembrane domains (Priori and Napolitano 2005). Forty percent of all RYR2 mutations are estimated to be de novo mutations. Only rare cases of rearrangements have been described and remarkably all concern RYR2 exon3 (Bhuiyan et al. 2007; Campbell et al. 2015). Since SCD can be the first manifestation of CPVT, genetic testing is clinically relevant, especially for the care management of family members.

At present, there is no genotype-based risk stratification or therapeutic approach in CPVT. Nearly one-third of possible or atypical LQTS cases with exertion-induced syncope have also been identified as positive for an RYR2 mutation, and accordingly, a clinical presentation of exercise-induced syncope and a QTc <460 ms should always first point to the consideration of CPVT rather than the so-called concealed or normal QT interval LQT1 (Medeiros-Domingo et al. 2009).

Since clinical overlap have been described between CPVT and cardiomyopathy cases (OMIM#604772), cardiomyopathy-related genes may be considered in case of negative diagnostic from the screening of the first and second gene panels (Fig. 13.3).

3.5 Cardiac Conduction Defect (CCD)

3.5.1 Clinical Description

Cardiac conduction defect (CCD) is a rare disorder associated with a risk of sudden death in the absence of pacing. Progressive cardiac conduction defect (PCCD), also called Lenegre-Lev disease, is the most frequent form, affecting the His-Purkinje pathway, and may progress to complete atrioventricular block (AVB) (Lev 1964; Lenegre 1964). Diagnosis relies on echocardiogram (ECG) findings showing an advanced conduction defect. Congenital structural CCD can be part of a syndrome and be associated with abnormalities in other organ systems or cardiac malformations including atrial septal defects, ventricular septal defects and tetralogy of Fallot (Baruteau et al. 2015).

Patients who receive pacemaker implantation appear to present an excellent prognosis except in those with LMNA mutations (see also below) which can lead to ventricular tachycardia and sudden cardiac death. In this population, ICD implantation is recommended in case of severe cardiac conduction defect (Anselme et al. 2013).

3.5.2 Genetic Testing

Transmission is usually autosomal dominant with incomplete penetrance and variable expressivity. Inherited forms of CCD are attributed to mutations in a wide variety of genes impacting the process of action potential generation and propagation. These electrical disturbances concern mainly isolated form of conduction defects and involve cardiac ion channel genes (SCN5A, HCN4, TRPM4, SCN1B) and cytoskeletal and inner nuclear membrane genes (LMNA). Mutations appear more frequently in SCN5A (Schott et al. 1999; Probst et al. 2003; Watanabe et al. 2008; Liu et al. 2010; Stallmeyer et al. 2012; Daumy et al. 2016). Point mutations and more rarely rearrangement in the gene encoding lamin A/C (LMNA) can lead to dilated cardiomyopathy associated with cardiac conduction system disease and arrhythmias and also to isolated CCD (Marsman et al. 2011). Genetic testing in suspected LMNA patients is of great interest since LMNA carriers present a high risk of SCD and a poor prognosis (Wolf and Berul 2006; Wolf et al. 2008). A mutation in the GJA5 gene encoding the gap junction Cx40 gene was identified in a family with progressive cardiac conduction defect (PCCD) characterized by AV block and wide QRS bundle branch block (Makita et al. 2012). Mutations in PRKAG2, which encodes for a regulatory subunit (γ-2) of adenosine monophosphate-activated protein kinase (AMPK), were found in patients with the Wolff-Parkinson-White (WPW) syndrome, a disease characterized by ventricular pre-excitation, atrial fibrillation and conduction defects such as sinoatrial and atrioventricular block and cardiac hypertrophy (Fig. 13.4; Table 13.4) (Wolf and Berul 2006). In the absence of mutations among the above candidate genes, the screening of the cardiomyopathy-associated genes might be considered (Sisakian 2014). Furthermore, in case of atrial septal defect (ASD), abnormalities in transcription-factor NKX2–5, GATA4 and TBX5 have been reported in congenital atrioventricular block (Schott et al. 1998; Garg et al. 2003; Basson et al. 1994).

Fig. 13.4
figure 4

Decision tree model for genetic testing in cardiac conduction defect

Full size image
Table 13.4 Genes identified in cardiac conduction defect (CCD)
Full size table

3.6 Arrhythmogenic Right Ventricular Cardiomyopathy/Dysplasia (ARVC/D)

3.6.1 Clinical Description

Arrhythmogenic right ventricular cardiomyopathy/dysplasia is a heritable cardiomyopathy characterized by life-threatening ventricular arrhythmias and progressive dystrophy of the ventricular myocardium with fibrofatty replacement. Generally referred to as a right ventricular disease, left or bi-ventricular forms prompted the use of the broader term arrhythmogenic cardiomyopathy (Bhonsale et al. 2015). The estimated prevalence in the general population is 1/2000–1/5000 with an age-related penetrance (usually adolescence-young adulthood) (Corrado et al. 2017). Moreover, ARVC/D more frequently affects males than females (up to 3:1) despite a similar prevalence of carrier status between sexes (Bhonsale et al. 2015). ARVC/D could account for up to 20% of cases of aborted cardiac arrest and SCD in the young, especially in athletes (Corrado et al. 2017).

Multiple criteria are required for diagnosis combining repolarization abnormalities, morpho-functional alterations, histopathological features on endomyocardial biopsy, family history and genetics, but diagnosis of ARVC/D is often complicated by its evolution over time (Marcus et al. 2010). Patients typically present with symptomatic ventricular arrhythmias, characterized by premature ventricular contraction (PVC) or VT with LBBB morphology and T-wave inversion in V1–V3 leads on basal electrocardiogram, leading to syncope or cardiac arrest. The first symptoms manifest before the age of 40 and are encountered in the large majority of patients (Orgeron and Calkins 2016).

3.6.2 Genetic Testing

ARVC/D is inherited predominantly as an autosomal dominant trait with incomplete penetrance and with a variable age-dependent expressivity. Inter- and intra-familial variation in disease severity and expressivity is frequently observed with coexistence of both classic RV and dominant LV pattern in the same family and/or life-threatening ventricular arrhythmias in probands vs. a more favourable prognosis in relatives (Sen-Chowdhry et al. 2007). In a few cases, ARVC/D inheritance follows an autosomal recessive trait (Naxos disease or Carvajal syndrome). These ARVC/D patients present extracardiac symptoms such as a cutaneous phenotype (palmoplantar keratoderma) and woolly hair (Corrado et al. 2017). ARVC/D is widely considered to be a desmosomal disease. Desmosomes are one of the junctional complexes at the intercalated disc that are essential for the structural and functional integrity of cardiac tissue. Desmosomal gene mutations lead to defective desmosomal components and to the development of structural abnormalities. Mutations in five desmosomal genes have been identified: plakophilin-2 (PKP2) desmoplakin (DSP), desmocollin-2 (DSC2), desmoglein-2 (DSG2), plakoglobin (JUP) and more recently the cadherin 2 (CDH2) (Mayosi et al. 2017). Non-desmosomal genes, transmembrane protein 43 (TMEM43), desmin (DES), lamin A/C (LMNA), titin (TTN), phospholamban (PLN) and alpha T-catenin (CTNNA3), have been associated with atypical forms of ARVC/D. Mutations in the regulatory region of transforming growth factor beta-3 (TGFB3) have also been reported, but their pathogenicity is still controversial. Rarely, mutations in the RYR2 gene and the SCN5A gene have been described (Table 13.5) (Te Riele et al. 2017). ARVC/D gene mutations are found in approximately 60% of ARVC/D index cases, of which the majority are desmosomal gene mutations (PKP2 10–45%, DSP 10–15%, DSG2 7–10%, DSC2 1–2%, JUP 1–2%) (Fig. 13.5), but the proportion of causal genes differs according to cohort location and ethnicity (Ohno 2016). Compound/digenic mutations may be found in 10–25% of patients and lead to earlier onset of symptoms, and carriers of mutations in DSP are more likely to present with left ventricular dysfunction, heart failure and SCD than carriers with mutations in PKP2 (Bhonsale et al. 2015; Rigato et al. 2013). Among single mutation carriers, premature truncating, splice-site and missense mutations were identified in, respectively, 60%, 23% and 14% of cases, but the outcomes did not differ significantly (Bhonsale et al. 2015). Entire PKP2 exons or even whole gene deletions have been recently described in families with a frequency of approximately 2% (Roberts et al. 2013; Li Mura et al. 2013).

Table 13.5 Genes identified in arrhythmogenic right ventricular dysplasia (ARVD)
Full size table
Fig. 13.5
figure 5

Decision tree model for genetic testing in arrhythmogenic right ventricular cardiomyopathy/dysplasia

Full size image

While the 2010 Diagnostic Task Force Criteria include identification of a putative ARVC/D susceptibility gene mutation as a new major diagnostic criterion, genetic testing should be used as a confirmatory tool for diagnosis of ARVC/D in index cases (Fig. 13.5). Presymptomatic screening is recommended in family members. If they carry the same mutation, they will then require clinical evaluation and long-term observation. Nevertheless, genetic findings require careful interpretation owing to the large number of genetic variants of uncertain significance (Andreasen et al. 2013).

3.7 Role of Genetic Testing in Sudden Unexplained Death (SUD)

3.7.1 Clinical Description

Sudden cardiac death is responsible for a large proportion of sudden unexpected deaths (SUD) in young individuals. A national study in Denmark estimated that SUD represents 29% of the deaths in persons aged 1–35 years after medicolegal investigation (Winkel et al. 2011). In 5–10% of all cases, SUD is caused by inherited cardiac diseases, especially cardiac channelopathies such as long QT syndrome, short QT syndrome, Brugada syndrome, catecholaminergic polymorphic ventricular tachycardia, progressive cardiac conduction disorder or early repolarization syndrome (Hayashi et al. 2009). In these cases, SUD may represent an early malignant arrhythmogenic presentation, preceding the development of the ECG indices. Therefore, when no identifiable ECG abnormalities can be observed or when only a post-mortem blood sample can be collected, genetic testing may provide insight into the causal mechanism leading to (fatal) arrhythmia. Comprehensive clinical investigation of SCD families would identify causes of death in 40% of cases, demonstrating that SCD were probably due to inherited heart disease (Tan et al. 2005; Behr et al. 2008). Next-generation sequencing (NGS), with the ability to screen hundreds of genes simultaneously, could provide confirmation of the clinical investigation or become the only possibility to point clinicians to a clinical diagnosis and identify the relatives at risk of similar severe arrhythmia. Furthermore, when patients present a borderline phenotype, genetic testing could encourage clinicians to further look for specific ECG abnormalities and employ adapted provocative tests.

3.7.2 Genetic Testing

The identification of the genes associated with cardiomyopathies and channelopathies has provided insights into disease mechanisms of cardiac arrest. Most genetic studies have focused on cardiac channelopathy-associated genes. Mutation analysis in 5 LQTS genes (KCNQ1, KCNH2, SCN5A, KCNE1, KCNE2) identified pathogenic variations in 11–17% of SCD cases (Chugh et al. 2003; Skinner et al. 2011; Winkel et al. 2012). More recently, post-mortem genetic testing for a large cohort of cases of autopsy-negative sudden unexplained deaths identified putative pathogenic variations in LQTS and CPVT genes in 20% of the cases (Tester et al. 2012). Another study on 34 genes of channelopathies identified 3 likely pathogenic variants in KCNH2, SNTA1 and RYR2 in 3 out of 15 (20%) SUD cases (Hertz et al. 2015). A prevalence of 5–10% of variants in the CPVT-associated gene RYR2 has been found in sudden infant death syndrome cases (Larsen et al. 2013). Exome analysis of 28 cases of autopsy-negative sudden unexplained death identified 3 rare variants in KCNQ1 and KCNH2 genes in 3 SUD cases (10%) and 4 rare variations in cardiomyopathy genes (DSP, MYBPC3 and TTN) (Bagnall et al. 2014). A recent screening of 100 genes in 61 cases of autopsy-negative sudden unexplained death (<50 years) identified likely pathogenic variants in 34% of the SCD cases, 40% associated with inherited cardiomyopathies and 60% associated with channelopathies (Christiansen et al. 2016).

However, the use of genetics as a clinical diagnostic tool must always be considered the context of the clinical case. Therefore, particularly in the context of SCD, genetic testing should complement or guide but never supplant the clinical investigation and segregation analyses. Without a clear clinical picture and a clear correlation between the disease and the genetic variants, causality may be nonconclusive, since several genes have been linked to several phenotypes or SCD could be caused by other unknown genetic causes or factors.

4 Genetic Testing and Technological Evolution: Progress and Side Effects

4.1 The Next-Generation Sequencing: High Throughput and Low Cost

As described above, the number of channelopathy-associated genes has increased dramatically in the last decades. With capillary sequencing, genetic testing was restricted to the major genes since this molecular analysis was expensive and had limited throughput. Since 2011, next-generation sequencing has revolutionized genetic diagnostics by offering the possibility to screen thousands of genes simultaneously coupled with a dramatic reduction (10,000-fold; www.genome.gov/sequencing costs) in sequencing cost (Kingsmore and Saunders 2011). However, knowing the number of genes to explore is crucial since one must maintain a good ratio in order to guarantee a reasonable cost (number of genes to be sequenced), data that are of sufficient quality to interpret the results (depth of sequencing) and a throughput that is compatible with an analysis of numerous samples. A minimum mean depth of 30X has been considered reasonable to achieve confident variant calls. In the absence of consensus gene design for channelopathies, a rationalized approach could consist in defining a set of genes comprising the channel subunits and partners predominantly implicated in rhythm disorders (see above). In case of negative results, a second panel that includes all genes associated with any primary rhythm disorders or even cardiomyopathies could be employed (Novelli et al. 2016). However, our broad sequencing capability could be moderated by interpretation, ethical and storage issues.

4.2 The More You Sequence, the More Complex Will Be the Interpretation

As illustrated by the genetics of the channelopathies, we observed a large genetic heterogeneity among each clinical entity coupled with gene overlap between them (Fig. 13.6). This gene overlap could also reveal a potential continuum among these channelopathies (Makita 2009). Moreover, several studies have reported channelopathies due to double mutations harboured by different genes (Kapplinger et al. 2010, 2009; Bauce et al. 2010). These observations suggest the need to screen large panels of genes. These panels are now widely employed within the molecular diagnostic centres covering between 50 and 200 candidate genes according technologies and diseases targeted. The continuous reduction of the sequencing cost tends to switch to whole exome (all coding regions of the genome) sequencing (WES), or even whole genome sequencing (WGS), to overcome the limitation of the panel(s). However, reagent cost for capturing approximately 100 genes remains 4 times lower than for WES and 10 times cheaper than for WGS. Aside from the financial point of view, drawbacks need to be anticipated before producing and analysing a large/whole gene scan. First of all, besides the promise that all coding regions and then the “Mendelianome” will be covered by a WES, daily practice reveals that WES actually provides a poorer coverage (% of targeted bases covered at a defined mean depth) than the targeted gene panels (Pua et al. 2016). Coverage for genetic diagnostic is crucial since all candidate DNA base pairs have to be confidently explored. Whole genome sequencing presents the advantage of uniformly covering the genome and facilitating genomic rearrangement identification (Sudmant et al. 2015) but at twice the WES price.

Fig. 13.6
figure 6

(a) SCN5A overlap syndromes and (b) genetic continuum among the cardiac channelopathies

Full size image

Today, genetic diagnostics face the challenge of variant interpretation uncovered by the large-scale sequencing technologies offering the advantage of simultaneously screening hundreds of candidate genes but at the same time revealing multiple rare genetic variations of unknown significance (Novelli et al. 2016). For example, rare genetic variations in associated channelopathy genes are also commonly found in the general population (Risgaard et al. 2013; Ghouse et al. 2017; Paludan-Müller et al. 2017). Large databases such as gnomad.broadinstitute.org consisting of >120,000 exomes and >15,000 whole genomes currently constitute a powerful tool to estimate the minor allele frequency (MAF) of a candidate variant. In addition, such databases furnish MAF in different broad categories of ethnicities which could help in the possible absence of demographically matched controls to prevent false-positive mutation. For this reason, numerous countries have developed their own national databases to reflect the specific genetic architecture of their population (Zawistowski et al. 2014). However, the rarity of a variant does not make it necessarily causal, and it can even turn out to be a variant of uncertain/unknown significance (VUS). Indeed, within an exome, between 300 and 600 rare variants will be identified which obviously do not at all cause a pathologic phenotype (Tennessen et al. 2012). Then, considering that 12,958 (64%) out of 20,344 genes in the genome (http://www.proteinatlas.org/) are expressed in the cardiac tissue, we can estimate that between 200 and 400 of them could be of potential interest in a context of a patient presenting a cardiac disease. This estimation puts into perspective the challenge of the number of variants generated by the NGS and our ability to provide a pertinent molecular diagnosis. Misinterpretation of a VUS can lead to inappropriate treatment such as an implantable cardioverter-defibrillator which could have a dramatic impact on the patient’s quality of life with possible severe complications and potentially serious psychologic consequences (Ackerman 2015).

This led to re-evaluation of the criteria defining the pathogenicity of a variant and then to modification of genetic test interpretations in order to avoid false positives (Ackerman 2015). Segregation studies among family members and/or functional experiments constitute the best approaches for causal mutation validation but remain rare owing to the time and cost investment required and are not adapted to a large number of variants. This sequencing technology revolution could therefore change the skill profiles required for genetic testing. High-throughput functional screening (in vitro) technologies to cope with massive VUS identification could emerge to characterize the variants isolated by data analysis and interpretations performed by bioinformatics. Indeed, after filtering out common variants by checking their MAF which is available in large public databases, additional tools could guide a geneticist to isolate the causal variant. The type of mutation can be considered (missense, nonsense, frameshift or splice-site variants) especially when loss of function of the protein can be predicted. Another argument of causality can reside into a previous observation of this variation in patients presenting the same phenotype and listed in dedicated databases (HGMD http://www.hgmd.cf.ac.uk/ac/index.php, CLINVAR www.ncbi.nlm.nih.gov/clinvar). When a variant is rare but never found in patients or in dedicated databases and no family members could have been tested, multiple algorithms can be used to indicate the likelihood of variant pathogenicity. These tools can combine information from nucleotide or/and amino acid conservation among species or include parameters such as modification of amino acid polarity or/and 3D protein structure. Others combine several algorithms to provide a consensus score (Table 13.6). Guidelines for interpretation and classification of genetic variations are proposed by the American College of Medical Genetics and Genomics to evaluate the pathogenicity of a variant (Richards et al. 2015).

Table 13.6 Non-exhaustive list of tools for variant consequence prediction
Full size table

This unexpected number of rare variants that harbours a genome and the difficulty of their interpretation could reinforce, especially for rare diseases such as channelopathies, the benefit of gathering clinical and genotype information in common databases to improve the knowledge acquired by the number of cases explored.

4.3 Genetic Testing and the Impact of Identifying a Mutation for Relatives and Predictive Testing

Because of the complexities involved, genetic testing should only be performed in dedicated specialized centres such as a cardiac genetics clinic or clinical genetics service where appropriate family management and genetic counselling before and after testing can be offered. This should be performed by cardiologists, nurses, clinical geneticists and genetic counsellors with specialized training in cardiovascular genetics.

Identification of a mutation in a family can often explain why the disease has occurred. It also allows for cascade testing of other affected and unaffected family members. Testing of affected family members is performed as a confirmation of their disease status and to exclude the possibility of a “phenocopy” (i.e. an individual who has an acquired rather than genetic cause of the same condition as other members of a family). This enables an accurate risk assessment for their offspring. Asymptomatic family members can be offered a predictive genetic test to clarify whether they are at risk of developing clinical disease and to determine the inheritance risk to their children. There may also be implications for participation in sports and employment.

4.4 Genetic Testing and Psychological Impact

The quality of life and psychological stress of cardiac variant carriers are most often not evaluated. A few studies have provided a quantification of the level of anxiety and depression in cardiac mutation carriers and their relatives (Hendriks et al. 2008; Christiaans et al. 2009). Genetic testing can help with anxiety regarding the disease especially when the clinical diagnostic is uncertain. Stress should also be considered in partners since the degree of anxiety is higher in carrier partners than in those of noncarriers. In this context, genetic counselling is essential to answer the questions generated by the clinical and genetic diagnosis to prevent anxiety. The uncertainty of the functional effect of VUS and, more generally, the large number of rare variants present in all individuals could also give rise to anxiety in genetic counsellors (Spoonamore and Ware 2016).

4.5 Large-Scale Sequencing and Incidental Findings

Aside from the complex VUS interpretation, known causal mutation may be uncovered and sometimes aside from the initial indication of cardiac diseases. These incidental findings are a matter of clinical and ethical debate since it is difficult to assess a clear benefit for a patient and/or a clinician to be informed of such mutations. The American College of Medical Genetics and Genomics (ACMG) recommends reporting variants belonging to a defined list of genes to an appropriate clinician for re-evaluation and surveillance of the patient and his family. The ACMG proposes a list of 57 genes associated with 24 syndromes, cancer, endocrinal or cardiac diseases. Of note, this recommendation concerns 9 channelopathy-associated genes and 11 cardiomyopathy genes (Green et al. 2013; Kalia et al. 2017). This risk should be well explained to the patient before performing a whole exome scan.

4.6 Large-Scale Sequencing Data, Storage and Privacy

The rapid development of sequencing throughput has encouraged us to screen increasingly larger numbers of genes and the entire genome in the very short term. Thus, there have been problems in bioinformatics to handle and characterize the great number of rare variants contained within the genome. Moreover, large-scale sequencing, in particular whole genome screening, also implies anticipating storage capability since data associated with the sequencing of a whole genome represents approximately 120GB. A 3-day run of a last-generation sequencer (i.e. Illumina X Ten solution) produces 16 whole genomes and 1.8 TB of data. For example, a centre such as UPPMAX accumulated 2200 TB of data from 2011 to the end of 2015 (Spjuth et al. 2016). In addition, whole genome data requires secure storage since they contain sensitive and private information. Aside from storage, privacy must be maintained during the phases of alignment, calling or comparison with public databases when they are not performed on a local level (Akgün et al. 2015). These large genome scans shake our habits and require thought before rushing into data production.

5 Conclusion

The last decade saw the new high-throughput sequencing technologies offering the opportunity to screen first a large panel of candidate genes, then all coding region and now the whole genome for a continuously decreasing cost. This allowed improving the efficiency of the genetic testing and consequently the care management for the patient and their family members. This opportunity to always screen more genes revealed also the high prevalence of rare variants present in the genome and therefore brings us to a new challenge consisting of isolate one(s) with an impact on the cardiac phenotype.