Article Text

Download PDFPDF

Sudden death and ion channel disease: pathophysiology and implications for management
  1. Rachel Bastiaenen,
  2. Elijah R Behr
  1. Department of Cardiovascular Sciences, St George's University of London, London, UK
  1. Correspondence to Elijah R Behr, St George's University of London, Division of Cardiovascular Sciences, Cranmer Terrace, London SW17 0RE, UK; ebehr{at}sgul.ac.uk

Abstract

The underlying aetiology of sudden arrhythmic death syndrome is predominantly inherited cardiac disease, and ‘channelopathies’ (cardiac ion channel disease) are the most common detectable cause of death. This heterogeneous group includes Brugada syndrome, long QT syndrome and catecholaminergic polymorphic ventricular tachycardia. Common features include variable penetrance, sudden death due to ventricular arrhythmias, and the absence of structural heart disease. The understanding of cardiac ion channel disease has been revolutionised by genetics. At present, genotype contributes to risk stratification in Brugada syndrome, long QT syndrome and catecholaminergic polymorphic ventricular tachycardia, and the future promises management tailored to the genetic diagnosis.

  • Sudden adult death syndrome
  • genetics
  • risk stratification
View Full Text

Statistics from Altmetric.com

Sudden arrhythmic death syndrome and ion channel disease

In the UK, sudden cardiac death (SCD) claims 400–800 young lives each year.1 A proportion of these deaths remain unexplained despite a careful and expert post mortem and toxicological analysis. This is known as the sudden arrhythmic death syndrome (SADS).2 3 Most cases are young men, who die during sleep or at rest, with only a minority reporting prior syncope. The underlying aetiology is predominantly inherited cardiac disease, and a history of other premature sudden deaths can be elicited in up to 30% of victims' families.3–5 Comprehensive cardiological evaluation can identify a diagnosis in up to half of SADS families, most commonly cardiac ion channel disease, also known as ‘channelopathies’.3 4 6 7 This heterogeneous group includes Brugada syndrome (BrS), long QT syndrome (LQTS) and catecholaminergic polymorphic ventricular tachycardia (CPVT). Other channelopathies outside the scope of this article are short QT syndrome, progressive cardiac conduction defect and inherited atrial fibrillation. Ion channel disease has also been implicated in the aetiology of sudden infant death syndrome, in ∼10% of cases.8

Channelopathies have features in common: a genetic basis; a structurally normal heart; and predisposition to life-threatening cardiac arrhythmias. There is genetic heterogeneity, meaning that many different genes may be causative (genotype). As a result of variable penetrance, members of a family group with the same genetic mutation may demonstrate different severity of clinical signs (phenotype), and it is possible for a mutation carrier to appear entirely unaffected. The basis for this genotype–phenotype discordance may be other genetic modifiers, but this has not yet been fully explained.9 Each member of a family group must therefore be assessed individually to determine their risk.

Genetic testing is predominantly used to confirm diagnosis by identifying a causative mutation in a patient with overt clinical phenotype. The chance of a positive result varies from 60–70% expected success rate in LQTS and CPVT to 20% in BrS.10 11 A negative result therefore does not rule out the condition. A positive genetic test can be used for cascade screening of asymptomatic family members. Less commonly, it may clarify diagnosis in a borderline case or form part of the molecular autopsy in SADS screening.4

Ion channels and the ventricular myocyte action potential

Ion channel disease is associated with mutations in genes encoding proteins that form or interact with the specialised channels that conduct sodium, potassium and/or calcium ions.12 These are pore-forming proteins, first proposed by Hodgkin and Huxley in their 1952 Nobel Prize theory of the nerve impulse and later confirmed by Neher and Sakmann using the patch clamp technique.13

Ion channel flux is essential for normal cardiac action potential initiation, propagation and repolarisation. The electrical impulse initiates in sinus node pacemaker cells and propagates via the specialised cardiac conduction system. This is possible because the cell membrane has an electrochemical gradient and myocytes are electrically coupled at their endplates by gap junctions. In the ventricular myocyte, the electrical potential at rest (phase 4) is negative (−90 mV) inside the cell with respect to the outside. The electrical gradient is maintained by the sodium calcium exchanger and inward rectifier currents (IK1 and IK2p).

During depolarisation (phase 0), the membrane becomes permeable to sodium ions via voltage-gated sodium channels (INa). These sodium channels are responsible for the amplitude and upstroke velocity of the action potential, important determinants for impulse propagation and conduction velocity.14 There is rapid inwards flux of sodium, and the electrochemical gradient reverses (+40 mV). Partial repolarisation begins when the Ito current is initiated (phase 1) and the membrane becomes permeable to potassium ions via potassium channels. These potassium channels control resting potentials, action potential waveforms, automaticity and refractory periods.14 During the plateau phase (phase 2), calcium ions enter the cell through voltage-gated L-type calcium channels (ICa). This triggers release of calcium from the sarcoplasmic reticulum via the ryanodine receptor voltage-gated ion channel (hRyR2), and activation of sarcomeric and myocardial contraction. Subsequently, there is increasingly rapid efflux of potassium via slow and rapid rectifier currents (IKs and IKr, respectively) with contributions from IKur, IKATP and IKACh. IKr is predominantly responsible for repolarisation (phase 3), which results in restoration of the negative electrochemical gradient. Simultaneously there is uptake of calcium ions by the sarcoplasmic reticulum, removal by the sodium calcium exchanger and cardiac relaxation (figure 1).

Figure 1

Ventricular myocyte action potential. The action potential of atrial and ventricular myocytes consists of five phases: depolarisation (phase 0), partial repolarisation (phase 1), plateau (phase 2), repolarisation (phase 3) and resting (phase 4). The long plateau phase and stable resting phase are characteristic of myocyte action potential. Specific ion flows contribute sequentially. The corresponding electrical activity as seen on the ECG is also displayed.

Disturbance of the balance of depolarisation and repolarisation is arrhythmogenic, and two main mechanisms are responsible for the generation of tachyarrhythmias in ion channel disease: triggered activity and re-entrant excitation. Triggered activity arises from afterdepolarisations which may be early or delayed. These are premature depolarisations that result from oscillations in intracellular calcium. If large enough, they can result in sufficient activation of voltage-gated sodium channels to cause sodium influx and a subsequent premature action potential that may result in an ectopic beat. In re-entrant excitation, there is spatial and temporal dispersion of refractoriness allowing re-entry and propagation of an action potential. Resultant ventricular tachycardia (VT) is usually polymorphic and can degenerate into ventricular fibrillation (VF).14

Brugada syndrome

The BrS is associated with right ventricular conduction delay and ST elevation in the right precordial ECG leads, syncope and sudden death due to VF.15 A type 1 Brugada ECG pattern describes coved ST segment and J point elevation ≥0.2 mV followed by a negative T wave (figure 2). BrS is diagnosed when the type 1 ECG is observed in two or more right precordial leads (V1–V3) in conjunction with other clinical features (table 1). The ECG abnormality may be concealed, dynamic and/or induced by sodium channel blocking agents, such as ajmaline and flecainide, which can be used to aid diagnosis.15 Placing right precordial leads in the second and third intercostal spaces increases sensitivity for detection of the type 1 pattern.15 16

Figure 2

ECG demonstrating the type1 Brugada pattern. There is characteristic coved ST segment elevation followed by T wave inversion in leads V1 and V2 (type 1 Brugada pattern) and ‘saddleback’ ST segment elevation in lead V3 (type 2 Brugada pattern).

Table 1

Diagnostic criteria for Brugada syndrome

Genetics

BrS is inherited as an autosomal dominant trait with incomplete penetrance. It has been associated with more than 250 mutations in seven different genes (table 2). Mutations in the SCN5A gene, encoding the α subunit of the Naν1.5 sodium channel, are identified in 21% of index cases (probands).10 In one series, loss-of-function mutations in genes CACNA1C, CACNB2 and CACNA2D1, encoding the L-type calcium ion channel, accounted for a further 12% of Brugada cases, some of which also exhibited a short QT interval.17 Mutations in the GPD1L channel interacting protein (ChIP), and SCN1B and SCN3B β subunits, have been identified in a few cases and result in loss of function with impaired INa. There is one example of a gain-of-function mutation in KCNE3, the putative β subunit for Ito. Recent evidence suggests that, in some families, the SCN5A variant represents a genetic modifier rather than a causative mutation.18

Table 2

Genetic mutations associated with Brugada syndrome

Pathophysiology

Studies in expression systems have demonstrated that loss-of-function mutations due to failure of channel expression (trafficking) or changes in functional properties (gating) result in reduced inwards depolarising sodium current INa.10 There are two competing hypotheses for the mechanism of ventricular arrhythmias. The repolarisation theory suggests that unopposed action of the Ito current in right ventricular epicardium leads to loss of the action potential dome. Dispersion of repolarisation allows re-entry during phase 2 of the action potential, providing the substrate for initiation of VT/VF. The depolarisation theory suggests that delayed conduction in the right ventricular outflow tract (RVOT) facilitates re-entry and subsequent arrhythmia.18

The majority of BrS remains genetically undefined, and there appears to be overlap with short QT syndrome and early repolarisation syndrome (see below). In addition, there is evidence that structural changes in the RVOT can result in the BrS phenotype. Therefore the concept of BrS as a single disease entity caused by reduced INa has been challenged. It is possible that, in different patients, the type 1 ECG pattern relates to different underlying mechanisms.18

Risk stratification and management

Arrhythmias usually occur at rest, particularly at night, and may be induced by febrile illness and sodium channel blocking agents. Management involves avoidance of drugs known to exacerbate the condition (see http://www.brugadadrugs.org) and use of paracetamol or ibuprofen to reduce fever. At present, the only reliable treatment to prevent SCD is the implantable cardioverter defibrillator (ICD). Isoproterenol is effective at suppressing VF in patients with BrS who present with electrical storm.19 Quinidine, a class Ia antiarrhythmic agent with Ito blocking effects, appears to suppress ventricular arrhythmias in high-risk patients.20 As yet there are insufficient prospective studies to consider quinidine first-line treatment, and a registry has been set up to address this. It can be used as adjunct therapy in patients with an ICD and recurrent appropriate shocks.20 Epicardial ablation of the RVOT has recently been shown to normalise the type 1 ECG pattern and prevent VT/VF recurrence in BrS.21 Catheter ablation may therefore prove to be an alternative strategy in patients with recurrent symptoms.

Risk stratification is controversial. There is general agreement that patients with aborted SCD or syncope are at high risk and should receive an ICD.15 However, the majority of sudden death victims have no recorded premorbid symptoms, and in asymptomatic patients with BrS the event rate appears to be lower than initially expected.22–25 Asymptomatic patients with spontaneous type 1 ECG patterns appear to be at higher risk.15 Although electrophysiological testing to induce sustained polymorphic VT/VF has been used to aid risk stratification, meta-analyses have failed to find an association between inducibility and risk. Programmed electrical stimulation in asymptomatic patients may, however, have a useful negative predictive value.26 Additional high-risk features in BrS include male gender and South East Asian ethnicity.15 Late potentials on signal-averaged ECG are more common in symptomatic patients but not an independent risk factor. QRS fragmentation appears to predict patients at risk of syncope due to VF.27

While patients with multiple SCN5A mutations present at a younger age than those with a single mutation, positive genotype alone is not an independent risk factor for SCD.10 One recent study has, however, suggested earlier and frequent recurrence of VF in symptomatic BrS patients with SCN5A mutations compared with those without.28 The severity of the biophysical consequence in INa resulting from an SCN5A mutation does appear to influence phenotype and possibly risk, with increased syncope and longer PR and QRS durations in patients with mutations leading to protein truncation compared with missense mutations.29 The functional impact of mutations may therefore support novel risk stratification strategies.

Early repolarisation syndrome

Early repolarisation (ER) was long considered a benign ECG variant, but in recent years has remerged as a marker of SCD. ER, particularly in the inferior and inferolateral leads, is associated with idiopathic VF. ER syndrome (ER with idiopathic VF) in one patient has been associated with a mutation in KCNJ8, and, in one series, mutations in CACNA1C, CACNB2 and CACNA2D1 were demonstrated in 16% of probands (as well as in 12% of BrS probands).17 There are similarities between ER and BrS, including response to heart rate and drugs. This has led to the suggestion that both form part of the same spectrum of disease, the ‘J wave syndromes’.30 However, important differences exist: ER changes are not provoked by sodium channel blocker administration and in ER syndrome the signal-averaged ECG rarely shows abnormal late potentials. Patients with ER and resuscitated SCD warrant ICD implantation and may benefit from quinidine therapy.31 At present, there are no risk stratifiers for asymptomatic subjects with ER.

The long QT syndrome

The LQTS is associated with QT prolongation, syncope and sudden death due to torsades de pointes (TdP) and VF.32 Diagnosis is made according to the Schwartz score (table 3).32

Table 3

Diagnostic criteria for long QT syndrome (the Schwartz score)

Genetics

The LQTS has both autosomal recessive (Jervell Lange-Neilsen) and dominant (Romano Ward) modes of inheritance but may also be polygenic. Although 12 different culpable genes have been identified (table 4), three subtypes account for 70–75% of definite congenital LQTS cases11: LQT1 results from loss-of-function mutations in KCNQ1 (encoding the α subunit of IKs), LQT2 from loss-of-function mutations in KCNH2 (encoding the α subunit of IKr), and LQT3 from gain-of-function mutations in SCN5A (encoding the α subunit of INa). The remaining nine genes include other loss-of-function mutations in potassium channels, gain-of-function mutations in calcium channels, and mutations in ChIPs such as ankyrin B and caveolin 3. Anderson–Tawil syndrome (LQT7) is a multi-system disorder associated with loss-of-function mutations in KCNJ2, encoding the IK1 current, expressed in both skeletal and cardiac muscle. The resulting phenotype consists of dysmorphology, hypokalaemic periodic paralysis and QT prolongation. Timothy syndrome (LQT8) is another multi-system disorder with gain-of-function mutations identified in CACNA1C encoding the α subunit of the L-type calcium current. The syndrome comprises QT prolongation with syndactyly, cognitive abnormalities, immune deficiency and intermittent hypoglycaemia. Compound or digenic disease occurs in ∼5% of cases.11

Table 4

Genetic mutations associated with congenital long QT syndrome

Pathophysiology

Different mutations affect ion channel function by different molecular mechanisms, but in all cases TdP results from triggered activity due to early afterdepolarisations.33 In LQT1, trafficking defects tend to result in a milder clinical phenotype (haploinsuffiency), but formation of dysfunctional ion channels from mutant and wild-type subunits causes more severe clinical disease (dominant negative effect).34 Trafficking defects are more common in LQT2, while LQT3 mutations result in late re-activation of sodium channels and increased late sodium current.

Risk stratification and management: role for genetics

Management of LQTS includes avoidance of QT-prolonging drugs (see http://www.azcert.org) and high-intensity sport. Medical therapy with β-blockers is first-line prophylaxis. Patients with aborted SCD or syncope and/or TdP despite β-blocker therapy should be considered for ICD implantation.35 Left cardiac sympathetic denervation is effective at reducing ventricular arrhythmias in patients intolerant of β-blockers, in children in whom an ICD may be avoided, and as an adjunct in those with recurrent ICD therapies.36 Permanent pacing with rate-smoothing algorithms can reduce bradycardia and pause dependent TdP, but does not reduce sudden death in high-risk patients, and therefore utilising the pacing function of an ICD is considered a safer alternative.

Risk assessment is predominantly based on clinical phenotype, the majority of data coming from the International LQTS Registry. The best predictors of events have been female adulthood, corrected QT interval (QTc) ≥500 ms, and previous syncope or aborted sudden death.37 A single episode of syncope has been associated with a sixfold increase in the risk of subsequent SCD.38 It has been proposed that patients presenting with aborted sudden death and spontaneous TdP be considered high risk, those with QTc ≥500 ms and previous syncope intermediate risk, and those with QTc ≤500 ms and no syncope low risk. Clinical risk assessment should be continuous, as there are age- and gender-specific differences. The risk of cardiac events is higher in males during childhood and females during adulthood, with gender risk reversal in the mid-teens.39 Arrhythmic episodes also appear to be associated with stressful life events.40

Genotype–phenotype studies have provided insights into ECG characteristics, clinical course and risk of cardiac events including sudden death (figure 3).41 Patients with LQT1 tend to have broad-based T waves and syncope or sudden death during exercise, particularly swimming. LQT2 patients have low-amplitude or notched T waves and syncope or sudden death with sudden auditory stimuli—for example, the morning alarm clock. Patients with LQT3 have long, flat ST segments, a tendency towards bradycardia and sudden death during sleep.

Figure 3

ECG characteristics in long QT syndrome subtypes. LQT1 tend to have broad-based T waves. LQT2 have low-amplitude or notched T waves. LQT3 patients often have long flat ST segments.

β-blocker therapy is most effective in LQT1 and LQT2. In LQT1, atenolol has been associated with 77% reduction in the risk of cardiac events in high-risk patients. In LQT2, nadolol was associated with 87% risk reduction.42 β-blocker therapy is least effective in LQT3.37 Sodium channel blockers, such as flecainide, have been shown experimentally to reduce the QT interval in some LQT3 patients, but may unmask or exacerbate ST elevation in those with Brugada overlap syndromes.43

The role of genotype in risk prediction is uncertain at present. A recent study of genetic carriers with a normal resting QTc showed a 10-fold increased risk of events compared with mutation negative relatives.44 In clearly affected individuals, it was thought initially that LQT1 and LQT2 patients were at higher risk of cardiac events, but events were more lethal in LQT3.45 Subsequent studies found a lower event rate in LQT1 and no significant difference between genotypes after adjustment for clinical risk factors.46 47 Exceptions are male children with LQT1 and female adults with LQT2, who appear to have elevated risk.38 For female adult LQT2 carriers, this risk appears greatest in the first year post partum.48 Transmembrane-located mutations and those with dominant negative effect are associated with increased risk independent of traditional risk factors in LQT1, and ion channel pore-forming mutations carry greater risk in LQT2.34 Future risk prediction may therefore be possible based on biophysical properties of mutated channels. In subjects with digenic or compound mutations, the QTc may be more prolonged than in monogenic carriers, and risk of cardiac events is higher. Common genetic variation also appears to act as an important genetic modifier. Single-nucleotide polymorphisms in NOS1AP, known to modulate the QT interval in the general population, have been associated with longer QTc intervals and more severe disease.49 In addition, incompletely penetrant LQTS may be more clearly expressed and individuals therefore more symptomatic when co-segregation of deleterious common polymorphisms occurs.50 These examples of oligogenic and polymorphic effects in LQTS demonstrate the concept of impaired ‘repolarisation reserve’, a reduction in the capacity of the heart to repolarise that may become more evident clinically when additional ‘hits’ to this reserve are present.

Catecholaminergic polymorphic ventricular tachycardia

CPVT was first described in children with syncope, sudden death, bidirectional or polymorphic VT and VF after a rapid increase in sympathetic tone, usually secondary to physical or emotional stress.51 Bidirectional VT shows an alternating QRS axis with 180° rotation on a beat-to-beat basis (figure 4).

Figure 4

ECG from an exercise tolerance test in a patient with catecholaminergic polymorphic ventricular tachycardia. Arrowed are multiple ventricular ectopics with bidirectional couplets. Note the alternating QRS axis with 180° rotation on a beat-to-beat basis.

Genetics

CPVT displays both autosomal dominant and autosomal recessive modes of inheritance with high penetrance. Approximately 55–65% of probands carry a mutation in the RYR2 gene, encoding the cardiac ryanodine receptor, which accounts for autosomal dominant disease.52 Mutations in the CASQ2 gene, encoding calsequestrin, a calcium-buffering protein, account for a small number of autosomal recessive cases. KCNJ2 mutations associated with Anderson syndrome may underlie a minority of cases.

Pathophysiology

Both the ryanodine receptor and calsequestrin are involved in intracellular calcium homoeostasis and excitation–contraction coupling. Calcium leakage from the sarcoplasmic reticulum is exacerbated during adrenergic stimulation, causing calcium overload and delayed afterdepolarisations. The mechanism of bidirectional VT is therefore triggered activity possibly arising from Purkinje tissue.53

Risk stratification and management

Patients should avoid competitive and high-intensity sports, stimulant drugs and digoxin. β-Blocker therapy is recommended, and the dose should be titrated against response using exercise testing to achieve optimal control of arrhythmias. Although early data suggested almost complete protection except in non-compliant patients, other studies observed recurrence of cardiac events even on maximal tolerated doses.51 54 ICD implantation is indicated after aborted SCD and for patients who have recurrence of syncope or VT despite β-blocker therapy. As in LQTS, left cardiac sympathetic denervation can be a helpful adjunct if symptoms persist.55 Blockade of L-type calcium channels by verapamil and of RyR2 channels by flecainide or propafenone appears to reduce intracellular calcium and in turn arrhythmia recurrence. Given in combination with β-blockers, these represent an alternative additional therapy.56

While evidence suggests that CPVT is a highly malignant condition, current data are limited and this view may change with accumulation of information from asymptomatic carriers. RYR2 mutations are, however, associated with earlier symptom onset, and male carriers have a fourfold higher risk of cardiac events.54 β-Blocker therapy is therefore recommended even in asymptomatic RYR2 mutation carriers.

Future directions

Improved risk stratification in the channelopathies is vital, particularly as ICD implantation is not without complication and our ability to discriminate those individuals likely to benefit is limited. Inappropriate therapy for sinus tachycardia and atrial arrhythmias is problematic and relevant in cardiac ion channel disease because patients tend to be young and physically active and SCN5A and RYR2 mutation carriers are prone to atrial arrhythmias.57 Long-term endovascular lead placement in young patients is also associated with inappropriate therapy secondary to lead failure requiring lead extraction, although this complication may be reduced in future by new ICD technologies with subcutaneous leads (S-ICD).57

Genotyping may be increasingly useful for risk stratification of patients. Rapid advances in sequencing technologies, together with cost reductions, have already brought genetic testing out of the research laboratory and into mainstream clinical practice. The number of known causative mutations increases steadily along with our understanding of the molecular mechanisms of arrhythmia initiation and modifiers of these events. This may fuel the development of ‘gene-specific’ therapy to directly counteract the functional consequences of mutations and improve management of affected individuals and their families. Translational collaboration between basic scientists and clinicians will be important to facilitate further advancement in this area.

Conclusions

Ion channel disease offers a paradigm for the understanding of a molecular lesion in the patient and its translation to phenotype and eventually management decisions. Nonetheless there are many gaps in our understanding, particularly of incomplete penetrance, risk stratification and the underlying pathophysiology of some of the conditions. Progress is rapid, and patients and their families will continue to benefit as our knowledge improves.

References

View Abstract

Footnotes

  • Funding Funds for research received from Biotronik and Boston Scientific.

  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.