The proband was a 65-year-old man who presented with an episode of polymorphic ventricular tachycardia after 3 days of treatment with a macrolide. The ECG after torsade de pointes showed a QTc interval of 600ms. A pedigree was taken in the patient interview and relatives were invited to undergo screening. The evaluation of the proband and his relatives included medical history, physical examination, 12-lead ECG, the Schwartz score,6 and 2-dimensional and Doppler echocardiography. Blood samples were taken for genetic analysis and the patient and his relatives gave written consent. The study was approved by the local ethics committee.

The QT interval was measured in II and V5 or V6, and was corrected by Bazett's formula (QTc = QT/RR0.5). We considered long QTc interval to be a value > 450ms in men or > 460 in women.7According to the recently published expert consensus on inherited primary arrhythmia syndromes,8 LQTS was diagnosed in the proband by the presence of a LQTS risk score ≥ 3.56 and a corrected QT interval ≥ 500ms in repeated ECGs. After the identification of the KCNH2-H562R mutation in the proband, the diagnosis in relatives was established based on the genetic study of the mutation.

Genomic DNA was extracted from peripheral blood samples using standard protocols. All exons of KCNQ1, KCNH2, SCN5A were amplified with intronic primers and sequenced in both directions using BigDye v1.1 chemistry in an ABI3130 analyzer (Applied Biosystems; Foster City, California, United States). The sequence for each gene was compared with the reference sequence in the National Center for Biotechnology Information sequence database. The pathogenicity of the new variant was evaluated using in silico software.

We used the SPSS statistical program version 15.0 for the analysis (SPSS; Chicago, Illinois, United States). Continuous variables were tested for normal distribution using the Kolmogorov-Smirnov test. Normally distributed variables are expressed as mean (standard deviation) and qualitative variables as absolute values and percentages. Mann-Whitney's U test was used to objectify the presence of differences between the carriers and noncarriers of KCNH2-H562R. The QTc shortening under treatment with beta-blockers was evaluated using the Wilcoxon test and the effect of different types of beta-blockers using the Kruskal-Wallis test. P values < .05 were considered statistically significant.

The index patient (II.2) was a 65-year-old Caucasian man that was admitted to our hospital because of a syncope episode while walking. The patient had been treated with clarithromycin for 3 days due to a respiratory infection. He reported several episodes of syncope since the age of 30 years, but had never been studied for this reason. A blood test showed a mild hypokalemia (K+ 3.1 mEq/L), with no other abnormalities. While he was at the emergency observation area, he experienced an episode of polymorphic ventricular tachycardia that degenerated into ventricular fibrillation and required electric cardioversion. The ECG postcardioversion revealed a QTc interval of 600ms (Figure 1A). Subsequently, clarithromycin was discontinued and the hypokalemia was corrected, as both produce abnormal prolongation of the QT interval. However, the QTc interval was still clearly prolonged (520ms), so treatment with bisoprolol was initiated.

Figure 1.

Twelve-lead electrocardiogram of the proband. A: post-cardioversion, showing a prolonged QTc interval of 560ms. B: after treatment with 5mg of bisoprolol/day, QTc interval of 490 ms.

(0.6MB).

Under treatment with a medium dose of bisoprolol, the QTc was reduced to 490ms (Figure 1B). However, the patient had symptomatic sinus bradycardia and multiple orthostatic presyncopal episodes. Thus, we decided to implant a DDD-R pacemaker (the patient refused an implantable cardioverter defibrillator) to increase the heart rate and the dose of beta-blockers, with the purpose of reducing the QTc interval and controlling the sinus bradycardia symptoms. After atrial pacing and maximum dose of bisoprolol, QT was reduced to 470ms. Subsequently, he has remained asymptomatic and had no arrhythmic events in the 24-h Holter ECG or the exercise test at annual follow-up visits during a 5-year follow-up period.

The patient's pedigree was taken and his first degree relatives were screened (Figure 2). Genetic analysis of cardiac ion channel genes was performed in the proband. This analysis revealed that the index case was heterozygous for the KCNH2-H562R missense mutation, being diagnosed with LQT2.

Figure 2.

Pedigree. Patients III.6, III.8 and III.9 died before reaching 1 year of age. Carriers of the mutation are marked with a (+). In blue are the carriers that remained asymptomatic despite a prolonged QTc interval, and in red the patients who presented symptoms (syncope, sudden death, aborted sudden death).

(0.14MB).

Twenty-four family members (Figure 2) were screened for the mutation previously found in the proband, and 12 relatives (50%) were carriers of the KCNH2-H562R mutation. Demographic characteristics, QTc intervals, and events in mutation carriers are shown in Table 1.

The family history revealed 3 unexplained deaths among children aged less than 1 year (III.6, III.8, III.9) (ECGs were not available and genetic study was not performed). Among the 13 carriers of the mutation, 6 patients (46%, including the index case) had cardiac symptoms (Table 2). Patient IV.4 had been diagnosed with epilepsy due to several nocturnal convulsive episodes. She showed a wide variation of the QTc interval in previous ECGs (QTc from 430ms to 480ms) and her cardiologist prescribed beta-blocker therapy, but she refused treatment. One night the phone rang while she was sleeping and she lost consciousness shortly after the conversation started; 6 years later, she died suddenly during sleep. Her father (III.2), who had to be an obligate carrier of the mutation, had died suddenly at home 15 years before at the age of 35 (no autopsy was performed). The daughter of patient IV.4 (patient V.1), had several episodes of syncope at rest despite treatment with propanolol (QTc 480ms; Figure 3A), requiring a left cardiac sympathetic denervation at the age of 6 years because she was considered too young for an implantable cardioverter defibrillator.

Figure 3.

Twelve-lead electrocardiogram of some KCNH2-H562R carriers. A: patient V.1 (QTc 480ms). B: patient IV.12 (QTc, 600 ms). Both reported several episodes of syncope at rest.

(0.97MB).

Currently, all living carriers (n = 12) are under treatment with beta-blockers, and only the aforementioned patient V.1 has shown cardiac symptoms during a 5-year follow-up period.

The typical notched T waves on the ECG were observed in 3 (23%) patients (Figure 3B), with broad T waves being the predominant morphology in the other patients. The QTc interval was longer in mutation carriers than in non-carrier relatives QTc, 493 [42] ms vs 418 [14] ms; P < .001). QTc was prolonged in symptomatic rather than in asymptomatic carriers, although the difference was not statistically significant (QTc, 517 [57] ms vs 477 [33] ms; P = .28) (Figure 4). QTc interval was normal in 23% of carriers (n = 3), with all of them being asymptomatic. Females had a higher QTc than males but this difference was not statistically significant (QTc, 508 [57] ms vs 481 [21] ms; P = .23).

Figure 4.

QTc interval distribution of mutation carriers according to their clinical status.

(0.05MB).

While the results of the genetic study were available, stress testing was done in those family members with a normal QTc to identify probable carriers. In the 3 individuals with normal QTc who were later found to be carriers of the mutation, the QTc interval showed similar behavior with exercise: during the first minute the QTc increased in 25 (5) ms, then progressively shortened during the rest of the exercise, and finally in the third minute of the recovery phase the QTc increased by 25 (10) ms. This pathological behavior allowed an early initiation of beta-blockers in these patients.

All carriers received beta-blockers (except IV.4, as noted above). Of these, 5 (42%) were initiated on bisoprolol, 4 (33%) on metoprolol, and 3 (25%) on propranolol (1 patient was excluded from the analysis due to ventricular pacing). The mean QTc shortening with beta-blockers was 50 (37) ms (n = 11; QTc, 493 [46] ms vs 442 [19] ms; P = .002). The QTc shortening was 44 (25) ms with bisorpolol, 47 (23) ms with metoprolol and 63 (73) ms with propranolol (P = .9).

The c.1685A>G, p.H562R (rs199472922) heterozygous variant was detected in exon 7 of the KCNH2 gene. This variant causes an amino acid change from histidine 562 to arginine, with this replacement taking place in the S5 transmembrane region of the protein (Figure 5). The amino acid involved is located in a highly conserved region between the species and paralogue genes. A bioinformatic study of the mutation was performed using “Mutation Taster”, resulting in a pathogenic variant with P = .99 (P indicates the prediction value; values close to 1 indicate high reliability of the prediction made).

Figure 5.

Location of the H562R missense mutation in the Kv11.1 channel.

(0.15MB).

In humans, KCNH2 is located on chromosome 7q35-36, and the coding region comprises 16 exons.2 The full length HERG1 subunit is composed of 1159 amino acids and has 6 transmembrane domains (S1–S6). The K+ selective pore is found between transmembrane domain S5 and S6 (amino acid residues 550 through to 650). The mutation that we found in the family studied is located in the S5 domain, being part of the channel pore. In 2002, Moss et al3 published the follow-up which extended over a period of 40 years analyzing 44 different HERG mutations, in which subjects with pore mutations had more severe clinical manifestations and a higher frequency of arrhythmia-related events (74% vs 35%; P < .001). Later, Shimizu et al4 proposed that the S5-loop-S6 region was also associated with the greatest risk in a study including 858 LQT2 patients with 162 different KCNH2 mutations. The reason for this is that pore mutations have a greater negative effect on cardiac delayed rectifier K+ current. However, only a small percentage of KCNH2 mutations have been characterized by a functional in vitro electrophysiological analysis.2,4

The KCNH2-H562R mutation was reported in 2 isolated patients in a study that examined the spectrum of mutations found in 2500 unrelated cases,9 but the authors did not describe the phenotype-genotype relationships of the genetic variants found. In addition, Sharma et al10 showed that the functional analysis of another mutation in the same amino acid (H562P) resulted in a marked reduction of the K+ current in a patient with several syncopal events. All these findings support the pathogenicity of the mutation found in the family described here.

However, assessment of the pathogenicity of an identified genetic variant is sometimes controversial, and is currently more common after the introduction of new genotyping technologies of next-generation sequencing.11,12 With these genotyping techniques, a large number of genes can be analyzed, thereby increasing the problems for interpretation. Thus, recent genome studies have questioned the pathogenicity of up to 17% of suggested diseases causing genetic variants in cardiomyopathies.11,12 In this article, we show how the clinical information and the analysis of cosegregation of the mutation with the disease allowed us to establish the diagnosis.

The ECG study showed a lower frequency of typical notched T waves,13 with only 23% of carriers showing this pattern. In addition, we observed a large variance in phenotype and QTc interval duration in family members sharing the same mutation. This variable penetrance could be partially explained by the coexistence of a single polymorphism on LQT-causing genes and/or unknown genes and/or multiple gene mutations.14

We observed no differences in the duration of the QTc in carriers with and without symptoms. Thus, we could not identify a priori by using the QTc interval value, those subjects who would develop symptoms during the follow-up. In addition, there were 3 asymptomatic carriers that showed a normal QTc. In this regard, it is known that there is an important overlap in the distribution of QTc between otherwise healthy subjects and patients with genetically confirmed LQTS. According to the published literature, 40% of the genotyped LQTS population exhibit QTc values < 460ms.15

Multiple studies have shown the usefulness of QTc response to exercise in patients with LQTS. Long QT syndrome type 1 patients usually have a marked prolongation of QTc during exercise, whereas LQT2 patients have a discreetly prolonged or not prolonged QT interval.16 Later, it was described that the QT interval is prolonged at 2-3min of the recovery phase in patients with LQT2,17,18 a finding also observed in carriers of this family.

Sex has been identified as an important independent factor affecting event risk in LQT2, wherein women display a significantly higher risk of cardiac events than men after the onset of adolescence.5 This is corroborated in our study, in which women had a higher QTc and 4 of the 6 symptomatic carriers (66%) were females.

In patients with LQT2, most cardiac events are associated with an increased sympathetic tone, emotional stress, and/or sudden auditory stimulation, which are the most common triggers for arrhythmia,19 such as in patient IV.4. A study published in 201020 showed that the most important risk factors for arousal-triggered cardiac events were female sex and pore-loop mutations, while those for exercise-triggered events were the presence of nonpore-loop mutations. However, other carriers of the mutation studied showed symptomsat rest. In this context, several studies have reported that about 30% of the cardiac events in LQT2 patients occurred at rest.21 Thus, REM sleep is associated with deep sympathetic activation, which could play an important role in triggering cardiac events in LQT2 patients.

It is not uncommon for LQTS patients to be diagnosed with epilepsy. Accordingly, the KCNH2 gene was initially discovered in the hippocampus, and its expression was later demonstrated in many regions of the central nervous system. In a study that included 343 LQTS probands, a personal history of seizures was more common in LQT2 than all other subtypes of LQTS combined.22 In the family studied, patient IV.4 had a previous history of pharmacologically treated epilepsy. Some antiepileptic drugs such as phenytoin and other antiepileptic drugs have arrhythmogenic potential because of their cardiac delayed rectifier K+ current blocker potential in vitro. Therefore, an ECG should be performed in all patients with syncope and seizures to rule out LQTS,2 since treatment with antiepileptic drugs may be inadequate and deleterious.

There are multiple causes of abnormal QT interval prolongation such as myocardial ischemia, cardiomyopathies, hypothermia, hypokalemia, hypocalcemia, hypomagnesemia, autonomic influences, the effects of drugs, and the recently described decrease in epicardial temperature.23 Hypokalemia and clarithromycin may have been the triggering factors in the index patient. Some studies have recently demonstrated a reduction in QTc interval in LQT2 patients with a combination of potassium supplementation and aldosterone-inhibition. We recommended treatment with potassium supplements in the proband and other carriers with hypokalemia or potassium levels toward the lower limit of normality. Furthermore, there is a wide array of drugs that prolong the QT interval and/or induce polymorphic ventricular tachycardia.24 In this context, prolonged QT interval and ventricular arrhythmias have been reported in patients treated with erythromycin because this drug produces an inhibition of the cardiac delayed rectifier K+ current channel.25 Similar to our index patient, several cases of QT prolongation and torsades de pointes after clarithromycin administration have been published. Since the structure of clarithromycin is similar to that of erythromycin, the arrhythmogenic properties should be similar in both.26 Other macrolides such as roxithromycin and azithromycin appear to be less arrhythmogenic in comparative in vitro studies.27

Therapy for LQTS is directed toward reducing the incidence of syncope and sudden death. The standard treatment includes beta-blockers, sometimes used in combination with an implantable cardioverter defibrillator.28 The effectiveness of beta-blockers was documented in a large study that included 869 LQTS patients, where cardiac events were reduced in probands from 0.97 to 0.31 and in affected family members from 0.26 to 0.15 events per year.29 In our cohort, with the exception of the girl V.1, we observed no events after the initiation of beta-blockers, observing a significant reduction in the mean QTc interval.

The protective effect of beta-blockers in the LQTS population is not uniform because its mechanism of action is probably related to the attenuation of adrenergic-mediated triggers. Prior data suggest that there is a high rate of cardiac events among patients with LQT2 and LQTS type 3 treated with beta-blockers.30 A later study published in 2011 suggests that within the LQT2 population there is a trigger-specific response to beta-blocker therapy, so that LQT2 patients experience a very low rate of exercise-triggered events during beta-blocker therapy, but have a high rate of arousal (more abrupt changes in heart rate than exercise) and nonexercise/nonarousal (not associated with sympathetic activation) triggered cardiac events despite recommended therapy.31 However, 13% and 17% of the patients who experienced a first arousal- and nonexercise/nonarousal-triggered event, respectively, experienced a subsequent exercise-triggered event.32 Migdalovich et al5 reported that medical therapy with beta-blockers was associated with a 61% reduction in the risk of sudden death in a population of 1166 LQT2 patients. A recent study comparing the efficacy of commonly used beta-blockers showed that propranolol had a higher QTc shortening effect than metoprolol and nadolol.32

The only patient that remained symptomatic underwent left cardiac sympathetic denervation. In a study involving 147 LQTS patients, left cardiac sympathetic denervation reduced cardiac events by approximately 90%.33

The most widely accepted risk stratification of LQTS is based on published mortality rates, and classifies patients in high (history of aborted and/or ECG-documented episodes of polymorphic ventricular tachycardia), intermediate (prior syncope and/or QTc > 500ms), and low (no prior syncope and QTc ≤ 500ms)29 risk. However, a recent study showed a different association between mutation characteristics and time-dependent differences in the clinical course of LQT2 patients. Therefore, after the onset of adolescence, women with and without high-risk mutations show increased risk for life-threatening cardiac events, whereas the risk in men only increased among carriers of the higher-risk pore-loop mutations.5 Therefore, the current trend is toward a combined assessment of clinical and mutation-specific data to improve risk stratification for life-threatening cardiac events in LQT2.