Author + information
- Received September 30, 2009
- Revision received December 3, 2009
- Accepted December 17, 2009
- Published online June 15, 2010.
- Marta Tomás, PhD*,†,
- Carlo Napolitano, MD, PhD*,‡,
- Luciana De Giuli, PhD*,
- Raffaella Bloise, MD*,
- Isaac Subirana, MS†,§,
- Alberto Malovini, MS∥,#,
- Riccardo Bellazzi, PhD∥,
- Dan E. Arking, PhD**,
- Eduardo Marban, MD, PhD‡‡,
- Aravinda Chakravarti, PhD**,
- Peter M. Spooner, PhD†† and
- Silvia G. Priori, MD, PhD*,‡,¶,* ()
- ↵*Reprint requests and correspondence:
Dr. Silvia G. Priori, Molecular Cardiology Laboratories, IRCCS Fondazione Salvatore Maugeri, Via Maugeri 10/10A, 27100 Pavia, Italy
Objectives We investigated the role of nitric oxide 1 adaptor protein (NOS1AP) as a genetic modifier of long QT syndrome (LQTS).
Background LQTS risk stratification is complicated by the phenotype variability that limits prediction of life-threatening arrhythmic events based on available metrics. Thus, the identification of new markers is desirable. Recent studies have shown that NOS1APvariations in the gene modulate QT interval in healthy and 1 LQTS kindred, and occurrence of cardiac events in healthy subjects.
Methods The study included 901 patients enrolled in a prospective LQTS registry. Three NOS1APmarker SNPs (rs4657139, rs16847548, and rs10494366) were genotyped to assess the effect of variant alleles on QTc and on the incidence of cardiac events. We quantified the association between variant alleles, QTc, and outcomes to assess whether NOS1APis a useful risk stratifier in LQTS.
Results Variant alleles tagged by SNPs rs4657139 and rs16847548 were associated with an average QTc prolongation of 7 and 8 ms, respectively (p < 0.05; p < 0.01); whereas rs4657139 and rs10494366 were associated with increased incidence of cardiac events (25.2% vs. 18.0%, p < 0.05 and 24.8% vs. 17.8% p < 0.05). Cox multivariate analysis identified rs10494366 minor allele as an independent prognostic marker among patients with QTc <500 ms (hazard ratio: 1.63; 95% confidence interval: 1.06 to 2.5; p < 0.05) but not in the entire cohort.
Conclusions Our results provide the first demonstration, to our knowledge, of a risk-conferring genetic modifier in a large LQTS cohort. Subject to confirmation in additional cohorts, we suggest that the NOS1APtag SNP genotype may provide an additional clinical dimension, which helps assess risk and choice of therapeutic strategies in LQTS.
Long QT syndrome (LQTS) is a genetic disorder caused by mutations that affect genes encoding ion-channel subunits or proteins that indirectly modulate the function of ion channels. The disease derives its name from the diagnostic electrocardiographic feature of prolonged repolarization (QT interval). The heart of LQTS patients is predisposed to develop life-threatening arrhythmias when exposed to physical or psychological stress (1). Beta-blockers are the mainstay therapy in LQTS (1). The use of the implantable cardioverter-defibrillator (ICD) is recommended in patients at high risk of sudden cardiac death (SCD) and in those who continue to have cardiac events despite antiadrenergic therapy (2). The identification of markers of increased arrhythmic risk is a major goal of epidemiological research in LQTS. In addition to extent of QT prolongation, the major prognostic indicators (3), sex and genetic locus, have also proven useful in predicting adverse events (4,5). More recently, the effort to account for the large phenotypic variability of the disease (6) has focused on the search for additional genetic modifiers. Polymorphisms at additional loci in LQTS genes have been proposed as factors that may modulate QT interval (7); however, conclusive evidence that 1 or more polymorphisms influence the phenotype, including risk of cardiac events, of the disease is lacking.
Recently, a genome-wide association study identified an association between single nucleotide polymorphisms (SNPs) in the nitric oxide 1 adaptor protein (NOS1AP) gene and QT interval (8). The QT-modifying influence of NOS1APSNPs was subsequently replicated in follow-up studies on the Framingham Heart and Rotterdam populations (9,10) as well as in other cohorts (11–16). Recently, the role of NOS1APgene variants in sudden cardiac death has been reported in the general population (17), but the role of variants within this gene in modulating susceptibility to cardiac events in LQTS patients is only beginning to be assessed (18,19).
Here, we assessed a large cohort of LQTS patients to address whether: 1) polymorphisms of the NOS1APgene modulate the duration of QT interval in LQTS; and 2) NOS1APSNPs are independent predictors of risk of cardiac events in LQTS patients and may therefore contribute to risk stratification.
From our clinical database and DNA bank of LQTS patients (4–6), we identified Caucasian individuals with a clinical and/or molecular diagnosis of Romano-Ward LQTS who had been screened for mutations in the KCNQ1, KCNH2, KCNE1, KCNE2, and SCN5Agenes (4–6) and for whom we had DNA and clinical data available on natural history, cardiac events, and response to therapy. Nine-hundred and one affected patients entered the study: 520 probands and 381 family members (216 first-degree relatives from 153 families). Families with more than 1 mutation were excluded.
Local and national laws regarding biomedical research, personal data protection, and confidentiality were respected. The study protocol was approved by the Maugeri Foundation Ethical Committee, and all participants or their legal guardians signed an informed consent.
QT interval was measured at the enrollment electrocardiogram (ECG). Heart rate QT correction was performed (QTc) in lead II (or lead I or III if it could not be measured in lead II) (4–6) from 12-lead ECGs with the use of Bazett's formula during stable heart rate (4–6). Based on previous studies (4,5), we used the QTc cutoff of ≥500 ms to define high-risk patients. Cardiac events were defined as syncope, torsade de pointes, cardiac arrest, or ICD shocks (4). As previously done in our studies (4–6), we used an observation period for cardiac events from birth to either attainment of age 40 years or the initiation of antiadrenergic therapy (beta-blockers or left cardiac sympathetic denervation). Symptomatic patients are considered those who experienced a cardiac event according to the preceding definitions.
On the basis of a recent report (17), we genotyped 3 polymorphisms (rs4657139, rs16847548, and rs10494366) located within the high linkage disequilibrium block in the 5′ region of the NOS1APgene (8); 3 SNPs (rs4657139, rs16847548, and rs10494366) showed minor allele frequency (MAF) >0.10 and were included in this analysis. SNPs and haplotypes are indicated by an uppercase “A” to refer to the “major” (common) allele and a lower case “a” to refer to the “minor” (rare) allele. We have compared whether the dominant, additive, or recessive model fitted better in order to predict QT or cardiac events. To do this, we performed linear and logistic regression for QT and cardiac events, respectively, taking the SNP as the predictor variable in a dominant (2 categories: AA vs. Aa or aa), additive (3 categories: AA vs. Aa vs. aa and treated as continuous variables), or recessive (2 categories: AA or Aa vs. aa). After estimating the model, the model with highest likelihood was chosen: the largest R2for linear model and the Nagelkerke's pseudo-R2for the logistic regression model. This analysis showed that the dominant model for QTc reached the best fit for both rs4657139 (p = 0.004) and rs16847548 (p = 0.008). For rs10494366, the linear trend was not significant for QTc (p = 0.209), indicating no significant effect of rs10494366 allele variation on QT, and both the additive and dominant model are equally appropriate. A dominant model was the best fit for cardiac events (p = 0.022). The SNPs having MAF >0.10 were included in the haplotypes in the following order: rs4657139 T>A, rs16847548 T>C, and rs10494366 T>G. The D′ parameter and the frequency of haplotypes ≥0.10, defined by the unphased genotypes and a dominant model, were estimated with the haplo.stats R package (Mayo Foundation for Medical Education and Research, Scottsdale, Arizona). This program uses the expectation-maximization algorithm to estimate the probability of all and of the most likely haplotype combinations. The inclusion cutoff for haplotypes was set to an estimated probability ≥0.80.
Genomic DNA was extracted from peripheral blood lymphocytes according to standard procedures. Methods for mutation screening on LQTS genes have been reported in detail previously (4–6).
Multiplex-PCR amplicons were purified with exonuclease I–shrimp alkaline phosphatase (USB, Cleveland, Ohio) and used as a template for fluorescent dideoxy single-base extension of unlabeled oligonucleotide primers with the ABI PRISM SNaPshot Multiplex Kit (Applied Biosystems, Foster City, California). A post-extension treatment with calf intestine alkaline phosphatase (New England BioLabs, Frankfurt, Germany), GeneScan-120 LIZ internal size standard addition, and heat denaturation were carried out before sequencing in an ABI PRISM 310 Genetic (Applied Biosystems).
Analysis of variance, paired and unpaired ttests, and cross-tabulations with Fisher exact test were used as appropriate. Classification tree analysis was used as an explorative unbiased technique to understand the structure of relationships between variables and to highlight cohort stratifications (20). We specifically used this approach in order to assess whether NOS1APSNPs might add information to risk stratification in our LQTS population when analyzed together with the known metrics (QTc, sex, and genotype). Classification trees were learned using the Orange open source data mining software (University of Minnesota, Duluth, Minnesota) with gain ratio as purity measure. Correction of QTc versus NOS1APstatistics for multiplicity of comparisons was performed by permutation test, and permutation-corrected p values were generated using an adaptive permutation approach as implemented in PLINK open source software (Massachusetts General Hospital Center for Human Genetic Research, Boston, Massachusetts). The cumulative probability of a first cardiac event before the age of 40 years and before antiadrenergic therapy was determined in the entire population and in each of the 3 most prevalent genetic loci with the use of the life-table method of Kaplan and Meier. Log-rank test with Bonferroni correction for multiplicity of comparisons was applied. Cox multivariate survivorship analyses were performed to evaluate the significance and independence of predictors of a first cardiac events or a first cardiac arrest alone. Values of p ≥ 0.05 were considered statistically significant. Unless otherwise specified, analyses were performed using SPSS software version 16 (SPSS Inc., Chicago, Illinois). Values of p < 0.05 were considered statistically significant. Means are reported ± SD.
The characteristics of the 901 LQTS patients (55% women, mean age at last follow-up of 31 ± 19 years) from 520 families included in the study are shown in Table 1.LQTS gene mutations were identified in 770 patients (85%): KCNQ1(n = 421; 55%), KCNH2(n = 243; 32%), SCN5A(n = 83; 11%), KCNE1(n = 21; 2.7%), and KCNE2(n = 2; 0.3%). In the remaining 131 patients, no mutations were found in the 5 genes. Two-hundred four patients experienced cardiac events.
MAF of SNPs and haplotypes
The MAF of the SNPs rs4657139 T>A, rs16847548 T>C, and rs10494366 T>G MAF were 0.42, 0.29, and 0.45, respectively, and these were in linkage disequilibrium (D′ >0.800). Out of the 901 subjects, NOS1APhaplotypes could be estimated with a probability ≥0.80 for 887 patients with 197 events (22.2%). The AAA, aaa, and aAa haplotypes were the only ones presenting with >0.10 frequency (0.51, 0.26, and 0.11, respectively). SNP MAFs and common haplotype frequencies are shown in Table 1. Haplotype analysis provided no information beyond what was indicated by SNP analysis alone; thus, these data are not included here. Detailed haplotype frequencies are reported in the online supplement (Online Table 1).
NOS1APand cardiac repolarization
SNP analysis showed that carriers (Aa or aa) of minor alleles tagged by SNPs rs4657139 and rs16847548 had a longer QTc of 7 ms (p = 0.047) and 8 ms (p = 0.009), respectively, in this cohort, whereas there was no difference in QT associated with the minor allele of rs10494366 (Table 2).
NOS1APand cardiac events
We assessed whether tag SNP variation in the 5′′NOS1APgene were associated with an increased risk of cardiac events in the absence of antiadrenergic therapy. The percentage of symptomatic patients was higher among rs4657139 rare allele carriers as compared with AA homozygotes (25.2% vs. 18.0%; sex-adjusted hazard ratio [HR]: 1.38, 95% confidence interval [CI]: 1.01 to 1.86; p = 0.040). Similarly, rs10494366 rare allele carriers were more symptomatic than AA homozygotes (24.8% vs. 17.8%; sex-adjusted HR: 1.42, 95% CI: 1.03 to 1.96; p = 0.031). Patients with the rs16847548 minor allele did not show an increase in risk of cardiac events. Similar univariate associations between SNPs and events were also observed with Kaplan-Meier survival analysis (Online Figs. 1 to 3).
QTc, sex, and genetic locus in LQTS
Cohort records were analyzed over an observation period of 23 ± 14 years, during which 204 patients (23%) experienced a cardiac event before 40 years of age or initiation of treatment with beta-blockers (Table 1). The mean age at first cardiac event was 14 ± 10 years. Syncope occurred in 178 of 204 (87%) patients, whereas cardiac arrest occurred in the remaining 26 individuals (13%).
Among LQT1, LQT2, and LQT3 patients (n = 747), Cox regression analysis showed that QTc, sex, and genetic locus were significantly associated with clinical outcome as previously reported (4,21). Symptomatic patients had a longer QTc (500 ± 48 ms, n = 156 vs. 466 ± 40 ms, n = 591; p < 0.0001; HR for QTc ≥500 ms: 2.81, 95% CI: 2.02 to 3.90; p < 0.0001) and were more likely to be female (68.5% females in the symptomatic group vs. 51.7% among asymptomatic cases; p < 0.0001; HR for females: 1.54, 95% CI: 1.02 to 2.16; p = 0.014). Genetic locus was also associated with the outcome since LQT2 and LQT3 patients had more cardiac events as compared with LQT1 cases (HR: 1.67, 95% CI: 1.20 to 2.32; p < 0.002).
NOS1APand outcome prediction in LQTS
We assessed the value of NOS1APgenotyping as a predictor of adverse outcomes and as a novel additional metric for risk stratification in LQTS. To this end, we first performed classification tree analysis for unbiased identification of risk clusters. The results showed that NOS1APgenotype becomes informative regarding event risk only in patients with QTc <500 ms, whereas QTc and sex dominated the model for patients with QTc ≥500 ms. Accordingly, Cox regression in the cohort of 901 patients showed that none of the NOS1APSNPs is a significant independent predictor of risk of events.
Overall, cardiac event distribution was remarkably different (p < 0.002) in the 4 subgroups of patients when further stratified by QTc and sex: females with QTc ≥500 ms were at highest risk of events (incidence 52%, n = 128), followed by a risk of 25% in males with QTc ≥500 ms (n = 77), a risk of 21% in females with QTc <500 ms (n = 370), and a risk of 13% among males with QTc <500 ms (n = 326). Kaplan-Meier analyses showed significant differences of event rates in the 4 groups (Fig. 1,Online Figs. 1 to 6).
When we focused on patients with QTc <500 ms, Cox regression analysis identified 2 independent markers of cardiac events: rs10494366 genotype (HR: 1.63; 95% CI: 1.06 to 2.50; p = 0.03) (Fig. 2)and female sex (HR: 1.61; 95% CI: 1.10 to 2.35; p = 0.01). Note that these data indicate female carriers of the minor allele of rs10494366 have a 2.6-fold increase in risk of arrhythmic events. Multiple regression analysis confirmed the lack of significant interaction between QTc and rs10994366. Interestingly, when we restricted the analysis to cardiac arrest only, there were 10 events in carriers of the rs10494366 minor allele and none in the common allele (log-rank p = 0.031) (Online Fig. 7).
NOS1APand LQTS mutation
Based on the published risk stratification schemes for LQT1, LQT2, and LQT3 patients (2,4), we investigated whether NOS1APSNPs can add new insights for risk stratification in this group of patients. Univariate analysis showed that in the 747 patients genotyped as LQT1, LQT2, or LQT3, variant alleles in rs10494366 and rs4657139 are significantly associated with risk of cardiac events (log-rank p = 0.016 and p = 0.037, respectively). Cox multivariate analysis demonstrated that: 1) the rs10494366 rare allele (HR: 1.47; 95% CI: 1.02 to 2.13; p = 0.04); 2) LQT2 or LQT3 genotype (HR: 1.76; 95% CI: 1.25 to 2.48; p = 0.001); 3) QTc ≥500 ms (HR: 2.78; 95% CI: 1.99 to 3.85; p < 0.0001); and 4) female sex (HR: 1.53; 95% CI: 1.09 to 2.16; p = 0.01) were each significant predictors of increased risk of cardiac events (Fig. 3).No significant multivariate associations were observed with rs4657139 rs16847548 (rs4657139 rare allele: HR: 1.27; 95% CI: 0.89 to 1.80; p = 0.18; rs16847548 rare allele: HR: 1.10; 95% CI: 0.89 to 1.51; p = 0.58).
Heritability of QT interval
Dissecting how genetic variation alters the electrical substrate has been the target of many investigations and an area of intense current research. Identifying the genes that modulate QT interval and their effect on the occurrence of cardiac events has, accordingly, become one of the primary goals of molecular studies of cardiac excitability. Several groups have performed candidate gene or genome-wide association studies (7,8,11,22–24) to identify genetic modifiers of repolarization. The candidate gene approach was initially used with the idea that variants of ion channels causing also the full-blown LQTS were potential modifiers. However, observations based on this hypothesis have been either inconsistent or remained unconfirmed (7,23–25). More recently, data from genome-wide association studies identified several variants of the NOS1APgene that play a major role in the modulation of cardiac repolarization (8,26,27). The link between NOS1APand QTc was provided first by Arking et al. (8), who identified SNPs associated with QT duration in the upstream 5′ regulatory region of NOS1AP. Such association came as a surprise since this gene had not been associated with cardiac electrogenesis, although it had been implicated in schizophrenia and dystrophic muscular atrophy (28,29). NOS1APencodes the CAPON protein, an adaptor molecule, which binds the neural nitric oxide (NO) synthase (Nos1) via a C-terminal PDZ domain, with scaffold and signaling ligands (e.g., dexras, dystrophins). Functional studies show that CAPON is a spatial controller of NO production, and it modulates action potential duration via an effect on Ca2+and K+currents (30,31).
Reproducible evidence obtained in a total of more than 15,000 subjects shows that not only is QT interval duration modulated by NOS1AP, but that naturally occurring polymorphisms in this gene exert the strongest influence of any of the 10 loci now known to modulate QT in humans (27). Very recently, Kao et al. (17) showed a positive association between some NOS1APSNPs with incidence of SCD in a large cohort of community SCD cases.
NOS1APDNA variants: from population studies to LQTS
Given this robust evidence for the role of NOS1AP, it becomes important to assess whether the same common variants account for the phenotypic variability among patients with LQTS (6,32), a disease in which cardiac repolarization is profoundly affected by deleterious mutations. A recent study has suggested a role for NOS1APin modulating LQTS phenotype in an ethnically isolated population carrier of A341Vmutation in the KCNQ1gene (19); however, it is unknown whether the result also applies to carriers of different mutations and, most importantly, whether NOS1APSNPs may be used for risk stratification. One major challenge for such studies in LQTS is the difficulty of collecting a cohort of adequate size for which clinical data and DNA samples are available. With this objective in mind, we queried the clinical database and DNA Biobank at the Maugeri Foundation and identified 901 patients in whom NOS1APpolymorphisms and changes in QTc and cardiac events could be studied. We chose to screen NOS1APpolymorphisms with a MAF higher than 10% among those that had been previously associated with QTc prolongation and/or increased risk of sudden death in different populations worldwide (8,17,26,27). We studied a population in the drug-free state in order to avoid the confusing effect of therapy on outcome; as a consequence, our results are pertinent to the role of NOS1APSNPs on the natural history of LQTS.
NOS1APrs4657139 and rs16847548 rare alleles were associated with a QTc prolongation of 7 and 8 ms, respectively, in this cohort. At first glance, such an effect seems small; however, it is within the range of the prolongation induced by drugs known to induce “acquired long QT syndrome,” such as terfenadine or cisapride, drugs that have been removed from the market by regulatory authorities based on an excess mortality related to an average QT prolongation of 6 ms (33). As shown by studies performed in different databases in the last decades, QT interval duration is the most powerful predictor of risk of arrhythmic events in LQTS (3,4). Thus, NOS1APcan represent a meaningful modifier in this disease.
NOS1APand risk of events in LQTS
To assess whether, besides prolonging the QT interval, NOS1APSNPs contribute to modulation of susceptibility to cardiac events, we first used classification tree analysis that showed a significant stratification role for NOS1APin patients with QTc <500 ms. Accordingly, Cox multivariate analysis showed that a QTc ≥500 ms is the strongest predictor of events followed by sex whereas NOS1APSNPs had a nonsignificant association in the entire 901 patient cohort.
On the other hand, in patients with QTc <500 ms (n = 696), we found that the minor NOS1APrs10494366 allele contributes quite significantly to risk stratification (HR: 2.6) (Fig. 2). Of note, not only the rate of cardiac events (syncope and cardiac arrest) but also the rate of cardiac arrests (hard end point) was significantly higher in the carriers of the minor allele (Online Fig. 7). Unfortunately, the total number of events was insufficient to reach multivariate statistical significance.
The new finding that variants in the NOS1APregion signal potential for increased risk of cardiac events and cardiac arrest in patients with QTc <500 ms has significant epidemiological and clinical implications as this group of patients represent between 65% and 75% of the total population (3,34). The more pronounced effect in patients with QTc <500 ms is conceivable since the risk of events is already so high in patients with longer QTc that it is unlikely that the presence of a single SNP can substantially modulate the risk.
The observation here that NOS1APrs10494366 minor allele has not been associated with risk of arrhythmias or sudden death in other investigations (17) is reminiscent of findings regarding positional variation in different individual NOS1APSNP associations reported in prior studies on other populations (8,10,11,26,27,35). As suggested by Kao et al. (17), genotyped SNPs may be considered proxy measures that are exposed to variable strength of the association with QT interval and SCD independently due to unequal linkage to the actual causal allelic variants. Therefore, we consider our finding on rs10494366 likely reflects population-specific (36) and/or disease-specific factors, as well as the possibility that there may be multiple functional elements within the NOS1APregion. This latter concept, if validated experimentally, would also be consistent with the presence of markers for increased risk of arrhythmic events independent from NOS1APeffects on QTc (17).
Based on the risk stratification model that we proposed in 2003 (4), which is now part of recommended practice guidelines for treatment of these patients (2), we analyzed our data to assess whether consideration of NOS1APvariation would provide additional strength to that model. As shown in Figure 3, genotyping variation in the NOS1AP gene has much potential to expand event risk stratification for many previously identified LQT1, LQT2, and LQT3 patients.
Although our findings will have to be replicated in an independent LQTS population before their clinical use is proposed, it is possible to comment on the potential role of NOS1APSNPs in treating LQTS patients.
When clinicians have to decide on therapy for LQTS patients, the dilemma of deciding whether the patient may need an implantable defibrillator, besides receiving beta-blockers, should be based on risk stratification schemes derived from validated epidemiological studies. We previously showed (4) that QT interval, sex, and genetic locus are the most important determinants of risk in LQT1, LQT2, and LQT3. Now, we provide evidence suggesting that NOS1APSNP genotype may be a novel metric, which could help to refine risk stratification. This approach has particular potential to identify individuals at higher risk among those with QTc <500 ms; until now, these patients, who represent the majority of LQTS clinical diagnoses, have been grouped in a lower-risk category. It is thus possible that the introduction of the screening for NOS1APSNPs may help the clinicians to identify subset of “QTc <500 ms patients” at higher risk, a supposition which will have to be validated by additional studies.
The potential of the discovery that NOS1APvariant genotypes may have validity towards this goal is illustrated in Figure 3. Based on the results of others, and our own (4) prior findings as extended here, we would consider that patients with a HR ≥6 be further evaluated as potential high-risk individuals who may benefit from ICD implantation. Patients with a HR between 4 and 6, depending on other clinical variables, may also be appropriate ICD candidates, whereas patients with a HR <4 in the absence of other risk factors (e.g., coronary or structural compromise) may be best treated with beta-blockers, again considering other cardiac risks and best clinical practices. Such considerations should obviously be in full compliance with clinical recommendations within the current American College of Cardiology, American Heart Association, and the European Society of Cardiology guidelines (2).
For a supplemental table and figures, please see the online version of this article.
This work was supported by Telethongrants nos. GGP04066and GGP06007, and by funds from the Ministero dell' Università e della Ricerca Scientifica e Tecnologica: FIRB RBNE01XMP4_006, RBLA035A4X_002, PRIN 2006055828_002(to Dr. Priori), the Alfonso Martín Escudero Foundation(to Dr. Tomás), the Donald W. Reynolds Foundation(to Drs. Chakravarti, Arking, Marban, and Spooner), and the Leducq Foundation(to Dr. Spooner). The first two authors contributed equally to this work.
- Abbreviations and Acronyms
- confidence interval
- hazard ratio
- implantable cardioverter-defibrillator
- long QT syndrome
- minor allele frequency
- nitric oxide 1 adaptor protein
- polymerase chain reaction
- sudden cardiac death
- single nucleotide polymorphism
- Received September 30, 2009.
- Revision received December 3, 2009.
- Accepted December 17, 2009.
- American College of Cardiology Foundation
- Priori S.G.,
- Napolitano C.,
- Schwartz P.J.,
- et al.
- Zipes D.P.,
- Camm A.J.,
- Borggrefe M.,
- et al.
- Moss A.J.,
- Schwartz P.J.,
- Crampton R.S.,
- et al.
- Bezzina C.R.,
- Verkerk A.O.,
- Busjahn A.,
- et al.
- Aarnoudse A.J.,
- Newton-Cheh C.,
- de Bakker P.I.,
- et al.
- Lehtinen A.B.,
- Newton-Cheh C.,
- Ziegler J.T.,
- et al.
- Tobin M.D.,
- Kahonen M.,
- Braund P.,
- et al.
- Eijgelsheim M.,
- Aarnoudse A.L.,
- Rivadeneira F.,
- et al.
- Kao W.H.,
- Arking D.E.,
- Post W.,
- et al.
- Schwartz P.J.,
- Monti M.C.,
- Insolia R.,
- et al.
- Breiman L.F.,
- Friedman J.H.,
- Olshen R.A.,
- Stone C.J.
- Sesti F.,
- Abbott G.W.,
- Wei J.,
- et al.
- Yang P.,
- Kanki H.,
- Drolet B.,
- et al.
- Pfeufer A.,
- Jalilzadeh S.,
- Perz S.,
- et al.
- Chang K.C.,
- Barth A.S.,
- Sasano T.,
- et al.
- Locati E.H.,
- Zareba W.,
- Moss A.J.,
- et al.
- Eijgelsheim M.,
- Newton-Cheh C.,
- Aarnoudse A.L.,
- et al.