Abstract
AIMS
Recent studies have shown that several genetic variants near the PITX2 locus on chromosome 4q25 are associated with atrial fibrillation (AF). However, the mechanism that mediates this association remains unclear. Basic murine studies suggest that reduced PITX2 expression is associated with left atrial dilatation. We sought to examine the association between single nucleotide polymorphisms (SNPs) near PITX2 and left atrial size in patients with AF.
METHODS
We prospectively enrolled 96 consecutive patients (mean age 60 ± 10 years, 72% male) with drug-resistant AF (57% paroxysmal, 38% persistent, and 5% long-standing persistent) who underwent catheter ablation. Following DNA extraction from blood obtained pre-operatively, SNPs rs10033464 and rs2200733 were genotyped using the Sequenom MassARRAY. Left atrial volume (LAV) was determined using three-dimensional imaging (CT or MRI prior to first ablation) and by investigators blinded to genotype results.
RESULTS
The minor allele frequencies at SNPs rs10033464 and rs2200733 were 0.14 and 0.25, respectively. Using multivariable linear regression, homozygosity for the minor allele at rs10033464 (recessive model) was independently associated with larger LAV (P = 0.002) after adjustment for age, gender, BMI, height, type, and duration of AF, left ventricular ejection fraction, history of hypertension, valve disease, and antiarrhythmic drug use. The strength of the association was reconfirmed in a bootstrap study with 1000 resamplings. In contrast, no association was found between rs2200733 variant alleles and LAV.
CONCLUSION
SNP rs10033464 near the PITX2 locus on 4q25 is associated with LAV. Left atrial dilatation may mediate the association of common variants at 4q25 with AF.
Keywords: left atrium volume, PITX2, atrial fibrillation, catheter ablation, single nuclear polymorphisms
Introduction
Atrial fibrillation (AF) is the most common cardiac arrhythmia in the US. Recent genome-wide association analyses have identified multiple loci that are associated with AF.1,2 One of the strongest associations with AF, confirmed by validation studies, involves two non-coding single nucleotide polymorphisms (SNPs) on chromosome 4q25 (rs2200733 and rs10033464).2 Polymorphisms at rs2200733 and rs10033464 are also associated with increased risk of AF recurrence after catheter ablation, elevated risk of cardioembolic stroke, and development of AF after coronary artery bypass grafting.3–5 Both SNPs are adjacent to the paired-like homeodomain transcription factor-2 (PITX2) gene, which has a major role in cardiac development.6 The PITX2 transcription factor has been linked to left atrial and pulmonary vein inflow morphogenesis.7 In addition, Chinchilla and colleagues have demonstrated that reduced PITX2 expression resulted in progressive enlargement of the murine left atrium (LA).8 However, no human studies have examined left atrial structural associations with SNPs near chromosome 4q25. In the present study, we tested the hypothesis that the 4q25 AF risk alleles are associated with left atrial volume (LAV).
Methods
Patients
The Institutional Review Board approved our protocol, and all patients provided written informed consent. The study complies with the principles of the Declaration of Helsinki. During the period from September 2009 to February 2011, we enrolled 100 consecutive patients with drug-resistant AF referred for catheter ablation at our institution. We excluded four patients from analysis because a three-dimensional image of the LA prior to their first ablation was unavailable, leaving 96 patients in the cohort.
Genotyping
DNA was extracted from blood samples using Qiagen’s Puregene DNA extraction kit. Genotyping of SNPs rs10033464 (HGVS name: g.111720761T>G) and rs2200733 (HGVS name: g.111710169C>T) was performed with the MassARRAY system of the Sequenom genotyping platform as part of a single multiplex SNP analysis (Sequenom). Each sample was tested at least three times, and at least two concordant results were required. Investigators were blinded to genotyping results.
Image acquisition
All individuals had either cardiac CT angiography (CTA) or MR angiography (MRA) to delineate end-diastolic left atrial dimensions. In patients who were undergoing their second or third ablation at the time of enrollment, we used CTA or MRA prior to the first ablation for LAV measurements. Of all patients, 81 (84%) underwent CTA and 15 (16%) underwent MRA prior to their ablation. CTA was performed at the end of the expiratory phase after intravenous administration of 70–80 mL of contrast (Isovue 370) using a multi-detector CT scanner (Toshiba’s Aquilion). MRA was performed after administration of 0.2 mmol/kg of gadopentetate dimeglumine (Magnevist; Bayer HealthCare) with a 1.5-T magnetic resonance scanner (Avanto; Siemens) and a body phased-array coil. All scans were ECG-gated and were acquired at atrial end-diastole, just before opening of the mitral valve. Of all scans, 24 (25%) were obtained during AF and the remaining scans were acquired during sinus rhythm. Our detailed image acquisition protocols for CTA and MRI have previously been reported.9
Measurement of LAV
A single experienced reader, who was blinded to patient genetic profile and demographic characteristics, performed image analysis. LAV was calculated using the prolate ellipsoid formula as previously described and validated.10 Using the endocardial border, we measured three orthogonal dimensions of the LA, including transverse (T), anterior–posterior (AP), and longitudinal (LO) diameters. Transverse diameter of the LA was defined as the distance between the midpoint of the right- and left-sided pulmonary veins in oblique axial images. The AP and LO diameters were measured at the midpoint of the transverse diameter in oblique axial and sagittal images (Fig. 1). These dimensions were then entered into the prolate ellipsoid formula: LAV = [(T × AP × LO) × (0.523)].
Figure 1.
The figure illustrates the methodology for measurement of LAV on a CT angiogram image. The transverse diameter of the LA (T) was defined as the distance between the midpoint of the right- and left-sided pulmonary veins in oblique axial images (A). The AP and LO left atrial diameters were measured at the midpoint of the transverse diameter in oblique axial (A) and sagittal images (B). These dimensions were then entered into the prolate ellipsoid formula: LAV = [(T × AP × LO) × (0.523)].
Statistical analysis
Continuous variables are presented as mean ± SD. Categorical variables are reported as frequencies and percentages. The association of LAV with 4q25 genetic variants was examined using multivariable linear regression models, adjusting for age, ethnicity, gender, BMI, height type and duration of AF, left ventricular ejection fraction, history of hypertension, antiarrhythmic drug use, and valve disease. Based upon univariate analyses of the association of LAV with 4q25 genetic variants, a recessive model was assumed for polymorphisms. To estimate uncertainty associated with the sample size, we utilized a nonparametric bootstrap study to confirm the significance of the associations. This method involves resampling observations with replacement from the data 1000 times, therefore, estimating the sampling distribution, standard error, and confidence intervals without assumptions about the distribution of the underlying population. Thus, the method is useful in situations with small sample size or when the theoretical distribution of the statistic is complicated. Using this method, we created new samples of 95 patients. We then executed the multivariable analysis, including those variables from model 3 that reached or approached statistical significance, and repeated the process for a total of 1000 times. These new samples were taken from the original data set using sampling with replacement, so it is not identical to the original sample.11,12 Results are reported as significant if P ≤ 0.05, and with Bonferroni correction for multiple comparisons, results should be considered significant at P ≤ 0.025. Statistical analyses were performed using STATA version 12.
Results
Of 96 patients in the cohort (mean age 60 ± 9.9 years, 72% male), 89 (93%) were Caucasians, 3 (3%) were African American and 4 (4%) were self-identified as other ethnicities. AF types were defined according to rhythm status prior to each patient’s first ablation procedure. Overall, 55 (57%) patients had paroxysmal AF, 36 (38%) had persistent AF, and 5 (5%) had long-standing persistent AF. The CHADS2 risk score was low (0 or 1) in 73 (76%) participants. Valve disease was present in six (6%) patients, including four with mild-to-moderate mitral regurgitation, one with severe mitral regurgitation, and one with mild aortic stenosis. There was a trend toward higher LA volume in patients with valve disease (143.7 ± 33.2 cm3 vs. 103.8 ± 5.4 cm3, P = 0.08).
The minor allele frequencies at SNPs rs10033464 and rs2200733 were 0.14 and 0.25, respectively. The genotype frequencies for rs10033464 and rs2200733 were in Hardy–Weinberg equilibrium (χ2 = 1.959, P = 0.162; χ2 = 0.474, P = 0.491, respectively). The r2 between rs1003464 and rs2200733 in our cohort was 0.003. In 44 (46%) patients, no risk variants at either rs10033464 or rs2200733 were observed. At least one risk allele at either rs10033464 or rs2200733 was detected in 12 (13%) or 31 (32%) patients, respectively. The remaining nine (9%) patients had at least one risk allele at both rs10033464 and rs2200733. The baseline characteristics of all participants and of participant groups by genotype have been summarized in Table 1.
Table 1.
Patients’ demographics (n = 96).
| VARIABLES | ALL PATIENTS (n = 96) | rs10033464 | P | rs 2200733 | P | ||||
|---|---|---|---|---|---|---|---|---|---|
| GG (n = 75) |
TG (n = 18) |
TT (n = 3) |
CC (n = 56) |
CT (n = 33) |
TT (n = 7) |
||||
| Age (mean ± SD), y | 60 ± 9.9 | 61 ± 9.4 | 58 ± 12.6 | 61 ± 6.5 | 0.8 | 62 ± 8.7 | 58 ± 12 | 60 ± 7.3 | 0.7 |
| Male, n (%) | 69 (72%) | 52 (69%) | 14 (78%) | 3 (100%) | 0.4 | 41 (73%) | 21 (63%) | 7 (100%) | 0.14 |
| Height, cm | 176 ± 10.5 | 176 ± 10.9 | 175 ± 10.1 | 178 ± 4.4 | 0.94 | 175 ± 10.9 | 176 ± 9.3 | 183 ± 10.7 | 0.16 |
| BMI, kg/m2 | 28.7 ± 5.4 | 29 ± 5.5 | 27 ± 4.6 | 31 ± 7.2 | 0.25 | 29 ± 5 | 29 ± 5.8 | 26 ± 6 | 0.45 |
| Atrial fibrillation duration until imaging | |||||||||
| <3 years, n (%) | 33 (34%) | 27 (36%) | 5 (28%) | 1 (33%) | 0.81 | 18 (32%) | 12 (36%) | 3 (43%) | 0.82 |
| ≥3 years, n (%) | 63 (66%) | 48 (64%) | 13 (72%) | 2 (67%) | 38 (68%) | 21 (64%) | 4 (57%) | ||
| Atrial fibrillation type | |||||||||
| Paroxysmal, n (%) | 55 (57%) | 41 (55%) | 14 (78%) | 0 | 0.83 | 31 (55%) | 20 (61%) | 4 (57%) | 0.9 |
| Persistent & long-standing persistent, n (%) | 41 (43%) | 34 (45%) | 4 (22%) | 3 (100%) | 25 (45%) | 13 (39%) | 3 (43%) | ||
| Congestive heart failure, n (%) | 12 (12.5%) | 9 (12%) | 2 (11%) | 1 (33%) | 0.54 | 8 (14%) | 3 (9%) | 1 (14%) | 0.76 |
| Valve Disease, n (%) | 6 (6%) | 5 (7%) | 0 (0%) | 1 (33%) | 0.08 | 4 (7%) | 2 (6%) | 0 (05) | 0.77 |
| LA Volume (cm3) | 106.3 ± 53.6 | 103.4 ± 53.0 | 101.7 ± 39.5 | 207.6 ± 59.2 | 0.003 | 107.4 ± 53.0 | 100.2 ± 57.3 | 126.6 ± 39.5 | 0.489 |
| Hypertension, n (%) | 53 (55%) | 42 (56%) | 10 (56%) | 1 (33%) | 0.74 | 26 (46%) | 24 (72%) | 3 (43%) | 0.043 |
| Diabetes mellitus, n (%) | 10 (10%) | 9 (12%) | 1 (5%) | 0 | 0.6 | 5 (9%) | 5 (15%) | 0 | 0.4 |
| Stroke/TIA, n (%) | 6 (6%) | 5 (6%) | 1 (5%) | 0 | 0.78 | 2 (4%) | 4 (12%) | 0 | 0.4 |
| Coronary artery disease, n (%) | 16 (17%) | 11 (14%) | 5 (28%) | 0 | 0.3 | 8 (14%) | 6 (18%) | 2 (28%) | 0.6 |
| Obstructive sleep apnea, n (%) | 18 (19%) | 14 (19%) | 3 (17%) | 1 (33%) | 0.79 | 11 (20%) | 6 (18%) | 1 (14%) | 0.94 |
| LVEF, (mean ± SD) | 55 ± 9 | 55 ± 9 | 60 ± 5 | 50 ± 15 | 0.58 | 55 ± 10 | 57 ± 6 | 52 ± 12 | 0.23 |
| Prior cardioversion, n (%) | 41 (43%) | 34 (45%) | 4 (22%) | 3 (100%) | 0.03 | 25 (45%) | 13 (39%) | 3 (43%) | 0.89 |
| Antiarrhythmic drug use prior to imaging, n (%) | 89 (93%) | 70 (93%) | 16 (89%) | 3 (100%) | 0.72 | 51 (91%) | 31 (94%) | 7 (100%) | 0.66 |
| CHADS2 Score, n (%) | |||||||||
| 0 | 37 (39%) | 27 (36%) | 8 (44%) | 2 (67%) | 0.89 | 25 (45%) | 8 (24%) | 4 (57%) | 0.55 |
| 1 | 36 (37%) | 29 (39%) | 7 (39%) | 0 | 22 (36%) | 14 (43%) | 2 (29%) | ||
| 2 | 16 (17%) | 14 (19%) | 1 (6%) | 1 (33%) | 6 (11%) | 9 (27%) | 1 (14%) | ||
| ≥3 | 7 (7%) | 5 (7%) | 2 (11%) | 0 | 5 (8%) | 2 (6%) | 0 | ||
Note: Values are presented in mean ± SD or n (%). Parentheses contain column percentages.
Abbreviation: LVEF, left ventricular ejection fraction.
The TT haplotype at rs10033464 was strongly associated with larger LAV (Table 2). In contrast, an association was not identified in those with the AF risk allele at rs2200733. In the unadjusted model, homozygosity at the T allele at rs10033464 was associated with 104.6 ± 29.7 cm3 increase in LAV (P = 0.001). To evaluate the effects of potential confounders, multivariable regression models with age, ethnicity, gender, AF duration, type of AF, BMI, height, left ventricular ejection fraction, history of valve disease, history of antiarrhythmic drug use, and history of hypertension were utilized. Model 1 controlled for demographic features such as age, ethnicity, and type of AF. As illustrated in Table 3, the magnitude of association for homozygosity at the T allele at rs10033464 with LAV was unchanged at 104.4 ± 31.6 cm3, and statistical significance was maintained (P = 0.001). In the fully adjusted model 2, the magnitude of association of homozygosity at the T allele at rs10033464 with LAV and statistical significance were maintained (102.7 ± 32.1, P = 0.0002, Table 3).
Table 2.
Association of LAV, CHADS2 score, and left ventricular ejection fraction with common variations on chromosome 4q25.
| SNPs VARIATIONS | rs10033464 | P | rs2200733 | P | ||
|---|---|---|---|---|---|---|
| (TT) | (TG OR GG) | (TT) | (CT OR CC) | |||
| LAV (cm3) | 207.6± 34.2 | 103.1 ± 5.2 | <0.001 | 126.6 ± 14.9 | 104.7 ± 5.8 | 0.30 |
| CHADS2 | ||||||
| 0 | 2 | 35 | 0.638 | 4 | 33 | 0.842 |
| 1 | 0 | 36 | 2 | 34 | ||
| 2 | 1 | 15 | 1 | 15 | ||
| 3 | 0 | 6 | 0 | 6 | ||
| 4 | 0 | 1 | 0 | 1 | ||
| LVEF (%) | 50.0 ± 8.7 | 55.9 ± 0.95 | 0.271 | 52.2 ± 4.8 | 55.9 ± 0.96 | 0.327 |
Note: Values are in cm3.
Abbreviations: LAV, left atrial volume; LVEF, left ventricular ejection fraction; SNPs, single nucleotide polymorphisms.
Table 3.
Magnitude of association between LAV and common variations on chromosome 4q25 in multivariable models adjusting for potential confounders.
| VARIABLE | UNIVARIATE | MODEL 1* | MODEL 2† | |||
|---|---|---|---|---|---|---|
| COEFFICIENT | P VALUE | COEFFICIENT | P VALUE | COEFFICIENT | P VALUE | |
| rs10033464 | 104.6 ± 29.7 cm3 | 0.001 | 104.4 ± 31.6 cm3 | 0.001 | 102.7 ± 32.1 cm3 | 0.002 |
| rs2200733 | 21.8 ± 21.0 cm3 | 0.303 | 21.6 ± 21.2 cm3 | 0.310 | 9.26 ± 23.0 cm3 | 0.689 |
Notes:
Model 1, adjusted for age, ethnicity, and type of AF.
Model 2, adjusted for age, ethnicity, AF duration, type of AF, BMI, height, gender, left ventricular ejection fraction, history of valve disease, history of antiarrhythmic drug use, and history of hypertension.
We further examined this finding by a nonparametric bootstrap method.11,12 This method confirmed that the magnitude of association of homozygosity for the T allele at rs10033464 with LAV withstands resampling (91.9 ± 38.2 cm3, P = 0.016, Table 4) and that the finding is unlikely to be incidental. Additionally, left ventricular ejection fraction (−1.3 ± 0.6 cm3 per 1% increase in ejection fraction, P = 0.026) and AF duration ≥3 years (22.0 ± 9.9 cm3, P = 0.038) were associated with LAV.
Table 4.
Magnitude of association between LAV and the rs10033464 SNP in a 1000-fold bootstrap simulation study.
| COEFFICIENT ± BOOTSTRAP STANDARD ERROR (CM3) | P VALUE | |
|---|---|---|
| rs10033464 | 91.9 ± 38.2 | 0.016 |
| Age (years) | 0.3 ± 0.5 | 0.607 |
| BMI (kg/m2) | 1.0 ± 0.9 | 0.266 |
| Left ventricular ejection fraction (%) | −1.3 ± 0.6 | 0.026 |
| Hypertension | −11.7 ± 11.4 | 0.306 |
| History of anti-arrhythmic drug use | −4.0 ± 16.8 | 0.810 |
| AF duration ≥3 years | 22.0 ± 9.9 | 0.027 |
Discussion
The main finding of this study is that homozygosity for the T allele at SNP rs10033464 is associated with larger LAV in models adjusting for potential confounders of left atrial size. Importantly, the association of homozygosity for the T allele at SNP rs10033464 with LA volume was independent of the association of left ventricular ejection fraction and AF duration with LA volume. To the best of our knowledge, this is the first report of an association between a SNP in close proximity to the PITX2 gene with left atrial structure in patients with AF. We also observed that variation at both rs10033464 and rs2200733 was common in our cohort of patients with drug-refractory AF referred for catheter ablation. In our cohort, 54% of patients had at least one of the risk variants. This prevalence is higher than the prevalence of 35% reported in a population sample of European descent.2
Left atrial dilatation can be the cause or the consequence of AF. It has previously been shown that left atrial size is independently associated with AF incidence.13 Increased LAV is also associated with decreased efficacy of catheter ablation for AF. Each 10 mL increase in volume is associated with a 14% increased risk of AF recurrence after ablation.14 However, the mechanisms that explain these associations are not yet well understood. Atrial dilatation and stretch may induce the activation of extracellular signal-related kinase Erk1/ERk2, leading to atrial fibrosis and thereby atrial arrhythmia.15 It appears, however, that atrial dilatation can increase AF susceptibility and perpetuation independent of variations in atrial fibrosis or tissue refractoriness.16 Conversely, it has also been shown that atrial dilatation may occur as a consequence of AF in patients without prior left atrial dilatation.17
The non-coding SNPs at rs10033464 and rs2200733 have been associated with AF, and are close to the PITX2 gene.2 The PITX2 gene product is a member of the pituitary homeobox family of transcription factors and appears to play an important role in embryologic morphogenesis.18 Chinchilla and colleagues have demonstrated that the expression of PITX2C (the predominant cardiac isoform of PITX2) is decreased in left atrial biopsies of patients with AF.19 Additionally, using a murine model, the same group demonstrated that reduced PITX2 expression in atrial myocardium resulted in enlarged atrial chambers.19 This effect of reduced PITX2 expression on chamber size is likely mediated in part by selective upregulation of Bmp10, a regulator of physiologic hypertrophy in the LA.19 Our findings support the findings of Chinchilla and colleagues in the murine model. After adjusting for potential confounders, variation at rs10033464 was associated with increased LAV in patients with drug-refractory AF referred for catheter ablation.
Study limitations
The study sample size is small and underpowered, and the findings may potentially be driven by false-positive, random findings. However, statistical analyses were driven by a specific hypothesis and limited to associations of polymorphisms at two specific sites with LAV. The majority of our patients were Caucasian; therefore, our results may not be generalizable to AF patients with other ethnici-ties. Conversely, the inhomogeneity of ethnicity may also bias the results; however, the reported associations persisted after adjustment for ethnicity in our multivariable model. Our cohort included patients with paroxysmal, persistent, and long-standing persistent AF; however, the associations persisted after adjustment for the type of AF. The LAV was measured using two modalities (CTA or MRA), which may introduce differential measurement error and/or variance. However, Wen and colleagues reported that measurements of LAV using these two modalities are very similar.20 In all patients with long-standing-persistent AF and 13 (36%) patients with persistent AF, images were obtained during AF. In comparison, patients with paroxysmal AF underwent imaging during sinus rhythm. In all cases, however, images were acquired just before the opening of the mitral valve, thus measuring maximum atrial volume. In addition, we controlled for type of AF in our multivariable models. Finally, no association was observed between polymorphism at rs2200733 and LAV. Previous reports have observed discrepant associations between the two SNPs and risk of AF in different populations21 and validation samples.2 These discrepant associations likely reflect variations in linkage disequilibrium of the two SNPs across different population samples. We implemented a resampling simulation study that confirmed the significance of the association; however, our results lack external validation/control and will need to be confirmed in future studies. We did not investigate the mechanism for the association of polymorphism at rs10033464 with atrial volume. However, the results are consistent with functional studies previously reported by Chinchilla and colleagues.19
Conclusion
Polymorphism at rs10033464, but not rs2200733, near the PITX2 locus on chromosome 4q25 is associated with increased LAV. Left atrial dilatation may mediate the association of common variants at 4q25 with AF. Additional studies with larger sample size are warranted to confirm these findings.
Footnotes
ACADEMIC EDITOR: Thomas E. Vanhecke, Editor in Chief
FUNDING: The research was funded in part by the Francis Chiaramonte Foundation to Dr. Calkins and National Heart, Lung, and Blood Institute Awards (K23HL089333 and R01HL116280) to Dr. Nazarian. The authors confirm that the funder had no influence over the study design, content of the article, or selection of this journal.
COMPETING INTERESTS: SN reports grants and personal fees from Biosense Webster, Inc. SN is a scientific advisor to Biosense Webster, and is the principal investigator for funding awarded to Johns Hopkins from Biosense Webster. HY reports cardiac arrhythmia research fellowship funding from the Norbert and Louise Grunwald fund. Other authors disclose no competing interests.
Paper subject to independent expert blind peer review by minimum of two reviewers. All editorial decisions made by independent academic editor. Upon submission manuscript was subject to anti-plagiarism scanning. Prior to publication all authors have given signed confirmation of agreement to article publication and compliance with all applicable ethical and legal requirements, including the accuracy of author and contributor information, disclosure of competing interests and funding sources, compliance with ethical requirements relating to human and animal study participants, and compliance with any copyright requirements of third parties. This journal is a member of the Committee on Publication Ethics (COPE).
Author Contributions
Conceived and designed the experiments: SN. Analyzed the data: YM, HY, SN, VZ. Wrote the first draft of the manuscript: HY. Contributed to the writing of the manuscript: all authors. Agree with manuscript results and conclusions: all authors. Jointly developed the structure and arguments for the paper: all authors. Made critical revisions and approved final version: all authors. All authors reviewed and approved of the final manuscript.
REFERENCES
- 1.Hokimoto S, Yasue H, Fujimoto K, et al. Expression of angiotensin-converting enzyme in remaining viable myocytes of human ventricles after myocardial infarction. Circulation. 1996;94:1513–8. doi: 10.1161/01.cir.94.7.1513. [DOI] [PubMed] [Google Scholar]
- 2.Perin EC, Silva GV, Sarmento-Leite R, et al. Assessing myocardial viability and infarct transmurality with left ventricular electromechanical mapping in patients with stable coronary artery disease: validation by delayed-enhancement magnetic resonance imaging. Circulation. 2002;106:957–61. doi: 10.1161/01.cir.0000026394.01888.18. [DOI] [PubMed] [Google Scholar]
- 3.Husser D, Adams V, Piorkowski C, Hindricks G, Bollmann A. Chromosome 4q25 variants and atrial fibrillation recurrence after catheter ablation. J Am Coll Cardiol. 2010;55:747–53. doi: 10.1016/j.jacc.2009.11.041. [DOI] [PubMed] [Google Scholar]
- 4.Wnuk M, Pera J, Jagiełła J, et al. The rs2200733 variant on chromosome 4q25 is a risk factor for cardioembolic stroke related to atrial fibrillation in Polish patients. Neurol Neurochir Pol. 2011;45:148–52. doi: 10.1016/s0028-3843(14)60026-8. [DOI] [PubMed] [Google Scholar]
- 5.Body SC, Collard CD, Shernan SK, et al. Variation in the 4q25 chromosomal locus predicts atrial fibrillation after coronary artery bypass graft surgery. Circ Cardiovasc Genet. 2009;2:499–506. doi: 10.1161/CIRCGENETICS.109.849075. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Burdine RD, Schier AF. Conserved and divergent mechanisms in left-right axis formation. Genes Dev. 2000;14:763–76. [PubMed] [Google Scholar]
- 7.Nazarian S, Hansford R, Roguin A, et al. A prospective evaluation of a protocol for magnetic resonance imaging of patients with implanted cardiac devices. Ann Intern Med. 2011;155:415–24. doi: 10.7326/0003-4819-155-7-201110040-00004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chinchilla A, Daimi H, Lozano-Velasco E, et al. PITX2 insufficiency leads to atrial electrical and structural remodeling linked to arrhythmogenesis. Circ Cardiovasc Genet. 2011;4:269–79. doi: 10.1161/CIRCGENETICS.110.958116. [DOI] [PubMed] [Google Scholar]
- 9.Tung R, Nakahara S, Ramirez R, Lai C, Fishbein MC, Shivkumar K. Distinguishing epicardial fat from scar: analysis of electrograms using high-density electroanatomic mapping in a novel porcine infarct model. Heart Rhythm. 2010;7:389–95. doi: 10.1016/j.hrthm.2009.11.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hof I, Arbab-Zadeh A, Dong J, Scherr D, Chilukuri K, Calkins H. Validation of a simplified method to determine left atrial volume by computed tomography in patients with atrial fibrillation. Am J Cardiol. 2008;102:1567–70. doi: 10.1016/j.amjcard.2008.07.048. [DOI] [PubMed] [Google Scholar]
- 11.Mooney CZ, Duval RD. Bootstrapping: A Nonparametric Approach to Statistical Inference. Newbury Park: SAGE Publications; 1993. [Google Scholar]
- 12.Baser O, Crown WH, Pollicino C. Guidelines for selecting among different types of bootstraps. Curr Med Res Opin. 2006;22:799–808. doi: 10.1185/030079906X100230. [DOI] [PubMed] [Google Scholar]
- 13.Benjamin EJ, Levy D, Vaziri SM, D’Agostino RB, Belanger AJ, Wolf PA. Independent risk factors for atrial fibrillation in a population-based cohort. The Framingham Heart Study. JAMA. 1994;271:840–4. [PubMed] [Google Scholar]
- 14.Shin SH, Park MY, Oh WJ, et al. Left atrial volume is a predictor of atrial fibrillation recurrence after catheter ablation. J Am Soc Echocardiogr. 2008;21:697–702. doi: 10.1016/j.echo.2007.10.022. [DOI] [PubMed] [Google Scholar]
- 15.Kawai N, Okada M, Suzuki A, et al. Noninvasive identification of the origin of ventricular premature contraction by phase analysis of radionuclide ventriculography. Kaku Igaku. 1984;21:1323–8. [PubMed] [Google Scholar]
- 16.Neuberger HR, Schotten U, Verheule S, et al. Development of a substrate of atrial fibrillation during chronic atrioventricular block in the goat. Circulation. 2005;111:30–7. doi: 10.1161/01.CIR.0000151517.43137.97. [DOI] [PubMed] [Google Scholar]
- 17.Sanfilippo AJ, Abascal VM, Sheehan M, et al. Atrial enlargement as a consequence of atrial fibrillation. A prospective echocardiographic study. Circulation. 1990;82:792–7. doi: 10.1161/01.cir.82.3.792. [DOI] [PubMed] [Google Scholar]
- 18.Poelmann RE, Jongbloed MR, Gittenberger-de Groot AC. Pitx2: a challenging teenager. Circ Res. 2008;102:749–51. doi: 10.1161/CIRCRESAHA.108.174847. [DOI] [PubMed] [Google Scholar]
- 19.Bertaglia E, Brandolino G, Zoppo F, Zerbo F, Pascotto P. Integration of three-dimensional left atrial magnetic resonance images into a real-time electroanatomic mapping system: validation of a registration method. Pacing Clin Electrophysiol. 2008;31:273–82. doi: 10.1111/j.1540-8159.2008.00986.x. [DOI] [PubMed] [Google Scholar]
- 20.Wen Z, Zhang Z, Yu W, Fan Z, Du J, Lv B. Assessing the left atrial phasic volume and function with dual-source CT: comparison with 3T MRI. Int J Cardiovasc Imaging. 2010;26(suppl 1):83–92. doi: 10.1007/s10554-009-9569-4. [DOI] [PubMed] [Google Scholar]
- 21.Lee KT, Yeh HY, Tung CP, et al. Association of RS2200733 but not RS10033464 on 4q25 with atrial fibrillation based on the recessive model in a Taiwanese population. Cardiology. 2010;116:151–6. doi: 10.1159/000318172. [DOI] [PubMed] [Google Scholar]