Long and Short QT Syndromes

Potassium-Related Long QT Syndrome: Most Common Forms

The delayed rectifier potassium current (I _K ) is a major determinant of phase 3 of the cardiac AP. It includes two independent components: one rapid (I _Kr ) and one slow (I _Ks ). The KCNQ1 and KCNE1 genes encode, respectively, the α-KvLQT1 and β-MinK subunits of the potassium channel conducting the I _Ks current. KCNQ1 mutations produce the LQT1 variant of LQTS, which is also the most prevalent.

Homozygous or compound heterozygous mutations in KCNQ1 and KCNE1 cause the recessive Jervell and Lange-Nielsen (JLN) variants of LQTS (JLN1 and JLN2, respectively). LQT5 is an uncommon variant (2%–3%) caused by mutations in the KCNE1 gene usually associated with low penetrance and a mild phenotype.

The KCNH2 and KCNE2 genes encode, respectively, the α-(hERG) and β-(MIRP) subunits of the potassium channel conducting the I _Kr current and are responsible for the LQT2 and LQT6 variants of LQTS, respectively. LQT2 is the second most prevalent variant of LQTS, accounting for 35% to 40% of LQTS genotype-positive patients.

Mutations in these genes causing LQTS provoke a loss of I _Ks or I _Kr current through different mechanisms. Defective proteins may coassemble with wild-type protein and exert a dominant-negative effect. Other mutations lead to defective proteins that do not assemble with wild-type peptides, resulting in a loss of function that reduces the I _Ks current by 50% or less (haploinsufficiency). Finally, defective peptides may not even reach the membrane of the cardiac cell because the mutations interfere with intracellular protein trafficking, this being the predominant mechanism in LQT2 mutations. Specifically, mutations in KCNH2 have been recently classified into four classes according to the biophysical property impaired: class 1, which involves disrupting the synthesis of K _V 11.1 α-subunits; class 2, which involves causing a trafficking defect; and classes 3 and 4, which involve affecting K _V 11.1 channel gating and permeation.

The I _Ks current is augmented by sympathetic activation; heart rate increases and is essential for QT adaptation. If I _Ks is defective, the QT interval will fail to shorten appropriately during tachycardia, thus creating a highly arrhythmogenic condition.

Other very rare genetic subtypes caused by mutations inducing a reduced potassium outward current are LQT11 (because of a mutation in the AKAP9 gene encoding for yotiao, a protein modulating I _Ks current), and LQT13 (because of a mutation in the KCNJ5 gene, causing a reduced acetylcholine-dependent inward rectifier current [I _KACh ]; see Table 96.1 ). ^, Andersen-Tawil (see Chapter 97 ) syndrome, caused by mutations in the KCNJ2 gene, was initially known as LQT7. Nevertheless, ATS should not be considered part of LQTS. It is a syndromic form characterized by periodic paralysis and distinctive dysmorphic facial or skeletal features, with prominent U waves but usually without a real QT prolongation. The arrhythmias of ATS range from premature ventricular beats (PVBs) to polymorphic/bidirectional ventricular tachycardia (VT); SCD could be somewhat more common than previously reported.

Sodium-Related Long QT Syndrome: Most Eclectic Forms

The SCN5A gene encodes the protein of the cardiac sodium channel. In vitro expression studies have shown that LQTS- SCN5A mutations produce the LQTS phenotype by inducing a gain of function, leading to an increase in the sodium inward current, which prolongs AP duration. The prevalence of LQT3 among genotype-positive LQTS patients is 10% to 15%. Other, less common genetic subtypes of LQTS are related to proteins that are part of the so-called sodium channel macromolecular complex. Specifically, mutations in ANK4 (LQT4), CAV3 (LQT9), SCN4B (LQT10), and SNTA1 (LQT12) have been associated with the disease, and they act by directly or indirectly increasing the sodium current (see Table 96.1 ). Mutations in SCN5A are also associated with overlapping phenotypes (LQTS, Brugada syndrome [see Chapter 95 ] and cardiac conduction defects [see Chapter 106 ]). The most common and best studied mutations associated with overlapping phenotypes are SCN5A -E1784K, 1795insD, ΔKPQ, and ΔK1500; it would appear that a negative shift of steady-state sodium channel inactivation and enhanced tonic block by class IC drugs represent the common biophysical mechanisms underlying the phenotypic overlap of LQT3 and Brugada syndrome.

Another peculiarity of sodium-related LQTS is that, despite being less frequent than potassium-related LQTS in adulthood, it constitutes the most common genetic form associated with LQTS-related perinatal mortality, including life-threatening arrhythmias in the first year of life, sudden infant death syndrome (SIDS), and sudden unexplained intrauterine fetal demise (IUFD; Fig. 96.1 ). ^, This opposite pattern between perinatal life and adulthood is likely because of a negative selection for SCN5A carriers in the perinatal and early infancy period.

The genetic variants identified on SCN5A in SIDS, IUFD, and neonatal LQT3 have different topologic distributions. Indeed, mutations located in the transmembrane and linker regions are significantly more often present in neonatal LQT3 cases than in SIDS/IUFD (86% vs. 22%; P = .002), whereas the variants located in the N-terminal and interdomain linkers (the regions where variants present in controls are more frequently observed) are found in 9.5% of neonatal LQT3 versus 44.4% of SIDS/IUFD ( P = .01). Furthermore, although none of the mutations identified in severe neonatal LQT3 cases is present in the publicly available control databases, 47% of functional variants associated with SIDS/IUFD are also observed in the general population. These considerations support the concept that severe neonatal LQT3 cases are caused by mutations able, per se, to cause a phenotype unlikely to be compatible with survival to adulthood. Indeed, the mutations identified in this subgroup of patients are mainly de novo. By contrast, the genetic variants identified in SIDS and IUFD cases have a wider range of clinical severity, and some of them, identified also in the general population, are probably not able to cause SCD, per se, although they can act as favoring factors. In further support of this view, we have observed that the SCN5A rare variants with a functional effect identified in SIDS and IUFD, also present in the publicly available databases, are significantly more prevalent in the SIDS/IUFD cohorts than in the general population (12 of 634 [1.9%] vs. 45 of 6261 [0.7%]; P = .0019).

Calcium-Related Long QT Syndrome: Most Complex and Dangerous Forms

In the heart, prolonged calcium current delays cardiomyocyte repolarization and increases the risk for arrhythmia. Five major genetic forms of LQTS linked to this mechanism have been described.

CALM1-3 (LQT14–16): calmodulinopathy

The small calcium-binding protein calmodulin (CaM) is an important modulator of different ion channels, such as the L-type calcium channel, the sodium channel, potassium channels, and ryanodine receptor. The evolutionary importance of this protein is highlighted by the presence of three different genes located in three different chromosomes (14q32.11, 2p21, and 19q13.32) encoding the same protein, a unique example in the human genome. Mutations in CaM genes can cause severe cases of LQTS, catecholaminergic polymorphic ventricular tachycardia (CPVT; see Chapter 91 ), or idiopathic ventricular fibrillation (IVF). In 2013 we reported on four infants and very young children presenting with major QT prolongation, recurrent CA, and poor response to therapy. Using exome sequencing in parents-child trios, we identified de novo mutations in either CALM1 or CALM2, two of the three human genes encoding CaM. Mutations in CALM3 were also identified. All mutation carriers were infants with recurrent VF usually triggered by sympathetic activation and major QT prolongation. The mutated proteins exhibit major reductions in calcium binding affinity, and their overexpression in heterologous systems shows that the LQTS phenotype is mainly because of a reduction in the calcium-dependent inactivation of L-type calcium channels. ^, Because this overexpression does not take into account the native allelic balance of CaM, in which six alleles are expressed, we generated patient-specific induced pluripotent stem cells (iPS) from one mutation carriers ( CALM1 -F142L) and differentiated them into cardiomyocytes. Our results confirm that the derangement of L-type Ca ²⁺ current (I _CaL ) inactivation is the dominant mutation consequence and suggest that the resulting repolarization delay is the most likely arrhythmogenic effect. The causes underlying this dominant negative mechanism have been recently discussed.

To better understand the natural history, clinical features, and response to therapy, we established an international calmodulinopathy registry that enrolled 74 subjects with a mutation in one of the CALM genes. The LQTS phenotype was the most prevalent and severe one. Indeed, life-threatening arrhythmias occurred in 78% of the cases with a median age of onset of 1.5 years, mean QTc was 594 ± 73 ms and a perinatal presentation occurred in 58%. Adrenergic stimulation was the major trigger for cardiac events and in the cases in which the arrhythmia onset was recorded, it was always an abrupt onset of VF. The ECG pattern was similar to those observed in LQT3 and Timothy syndrome, with late-onset peaked T waves. Perinatal presentation was often associated with an extremely long QTc (628 ± 62 ms), 2:1 AV block, and T wave alternans. The mutations identified were mainly de novo with a few cases of mosaicism identified ; therefore no true familial cases were so far detected. β-Blocker monotherapy was insufficient to protect these patients and a combination of β-blockers, mexiletine, and LCSD was the best combination to improve protection. Nonetheless, most patients needed an ICD to prevent SCD ( Fig. 96.2 ). CALM-LQTS might not always be a pure electrical disorder because some patients had neurologic features unrelated to cardiac arrest, and others had cardiac structural abnormalities.

Genome-Wide Association Studies in Long QT Syndrome Patients: Is It Always a Monogenic Mendelian Disorder?

A genome-wide association study (GWAS) was recently conducted in 1656 unrelated LQTS patients of European or Japanese ancestry and in 9890 controls to evaluate the influence of single nucleotide variants (SNVs) in disease susceptibility. Most cases (89%) were genotype positive (800 LQT1, 661 LQT2, and 123 LQT3), whereas 197 were genotype negative. Three loci reached the threshold for genome-wide statistical significance (rs12143842 on NOS1AP, rs179405 on KCNQ1 , and rs179405 on KLF12 ) in both the European and Japanese populations; each of these three SNVs had been previously associated with QT interval duration in the general population. Furthermore, KCNE1 -D85N showed an association at the suggestive threshold of P > 10 ^-6 . Subsequently, the aggregate effect of the 68 QT-SNVs that had shown association with the QT interval in the general population (i.e., the weighted QT polygenic risk score [PRS _QT ]), was evaluated and showed a significant association with LQTS diagnosis in both populations. Interestingly, the association with PRS _QT was greater in genotype negative patients compared with genotype positive patients, suggesting a polygenic architecture in genotype negative LQTS cases.