A total of 3,056 consecutive patients who underwent radiofrequency catheter ablation for drug-refractory AF from June 1998 to June 2016 at two institutions (Korea University Anam Hospital and Gachon University Gil Medical Center) were screened for this study. Inclusion criteria consisted of patients having early-onset AF, which was defined as an arrhythmia occurring before the age of 40 years. Exclusion criteria were as follows: 1) patients with structural heart disease, 2) patients with previous history of ischemic heart disease. Ischemic heart disease was diagnosed based on functional or anatomical evaluation referring to patients' symptom as the guideline recommendation.9 Among them, 89 patients (2.91%) met the inclusion and exclusion criteria. These 89 patients were considered for the present analysis and constituted the study population. The clinical and electrophysiological characteristics of these patients, as well as the clinical outcomes of catheter ablation were subsequently reviewed.

Sixteen SNPs including rs13376333, rs10465885, rs10033464, rs2200733, rs17042171, rs6843082, rs7193343, rs2106261, rs17570669, rs853445, rs11708996, rs6800541, rs251253, rs3807989, rs11047543, and rs3825214 reported in the literature to be associated with AF were genotyped in this study.31011 DNA was extracted from the buffy coat fraction of blood samples taken from the 89 patients. Genotypes were screened using single base primer extension assay using the ABI PRISM SNaPShot Multiplex Kit (Applied Biosystems, Inc., Foster City, CA, USA) according to manufacturer's instructions. Briefly, the genomic DNA flanking the interested SNP was amplified by polymerase chain reaction (PCR) with forward and reverse primer pairs and standard PCR reagents in a 10 μL reaction volume, containing 10 ng of genomic DNA, 0.5 pM of each oligonucleotide primer, 1 μL of 10× PCR buffer, 250 μM dNTP (2.5 mM each), and 0.25 U DiaStar Taq DNA Polymerase (5 U/µL) (SolGent Co., Ltd., Daejeon, Korea). The PCR reactions were carried out as follows: 10 minutes at 95°C for 1 cycle, and 35 cycles of 95°C for 30 seconds, 60°C for 1 minutes, 72°C for 1 minute, followed by 1 cycle of 72°C for 10 minutes. After amplification, the PCR products were treated with 1 U each of shrimp alkaline phosphatase (SAP) (USB Corporation, Cleveland, OH, USA) and exonuclease I (USB Corporation) at 37°C for 75 minutes and 72°C for 15 minutes to purify the amplified products. One microliter of the purified amplification product was added to a SNaPshot Multiplex Ready reaction mixture containing 0.15 pmol of genotyping primer for the primer extension reaction. The primer extension reaction was carried out for 25 cycles of 96°C for 10 seconds, 50°C for 5 seconds, and 60°C for 30 seconds. The reaction products were treated with 1 U SAP at 37°C for 1 hour and 72°C for 15 minutes to remove excess fluorescent dye terminators. One microliter of the final reaction samples containing the extension products were added to 9 μL of Hi-Di formamide (Applied Biosystems, Inc.). The mixture was incubated at 95°C for 5 minutes, followed by 5 minutes on ice, and then analyzed by electrophoresis in an ABI Prism 3730xl DNA analyzer. Analysis was carried out using Genemapper software (version 4.0; Thermo Fisher Scientific, Waltham, MA, USA).

Anti-arrhythmic medications were discontinued at least five half-lives before the procedure. Amiodarone was discontinued at least one month before the ablation procedure. Transesophageal echocardiography was performed within 24 hours before the procedure to exclude the presence of left atrial (LA) thrombus. The ablation procedure was performed under sedation with intravenous propofol under continuous monitoring of blood pressure and oxygen saturation. The high right atrium (RA), low RA, and coronary sinus were mapped with a decapolar catheter and steerable duo-decapolar catheter, respectively, inserted through the left femoral vein. A quadripolar catheter was also placed in the superior vena cava. Intracardiac electrograms were recorded using an electrophysiology system (Prucka CardioLab^™ General Electric Health Care System Inc., Milwaukee, WI, USA). After the double trans-septal puncture, anticoagulation was started with unfractionated heparin, maintaining an activated clotting time between 300 and 350 seconds. We used 3-dimensional (3D) mapping guided geometry (NavX System, Abbott, CA, USA) for electroanatomical mapping in all patients. To avoid or minimize applications of radiofrequency energy near the esophagus, the esophagus was visualized by barium swallow before sedation.

The stepwise approach for ablation was performed under the guidance of fluoroscopic and 3D mapping. The AF triggering pulmonary vein (PV) and non-PV foci were evaluated under high dose isoproterenol infusion (10–20 μg/min) in cases of paroxysmal AF at the beginning of the procedure. All patients initially underwent circumferential antral ablation with the endpoint being the electrical PV exit and entrance block or dissociation. Paroxysmal AF patients underwent additional ablations depending on the presence of AF triggering non-PV foci or inducibility test results. If AF was sustained following antral ablation of the PVs in persistent AF patients, further ablation was guided by automated complex fractionated atrial electrogram (CFAE) maps of the LA and then the RA, which were defined previously.12 The endpoints of CFAE-guided ablation were a significant reduction in the CFAE amplitude (> 80%), electrical silence, organized atrial tachycardia (AT), or the termination of AF. CFAEs of the RA were targeted if AF persisted after extensive LA ablation. Linear ablation at the cavotricuspid isthmus was performed in all persistent AF patients either before or after restoring the sinus rhythm and bidirectional conduction block was confirmed. When AF converted to AT, activation mapping and ablation for AT were performed until a sinus rhythm was restored. If a patient had more than one stable AT, an attempt was made to map and ablate all ATs. The endpoints of catheter ablation were the absence of AF-triggering foci or non-inducibility in paroxysmal AF patients and the termination of AF or AT13 and then non-inducibility of AT (cycle length > 280 ms) by burst atrial pacing in persistent AF patients. If AF or AT did not terminate after ablating all target sites, the sinus rhythm was restored by electrical cardioversion. Radiofrequency ablation was delivered at a target temperature of 48°C and power in the range of 25–35 W (Stockert generator; Biosense Webster, Inc., Diamond Bar, CA, USA or IBI 1500T11; NavX System, Abbott, CA, USA) using a 4 mm open irrigated-tip catheter (Thermocool; Biosense Webster, Inc. or Cool Path Duo; NavX System).

Patients were monitored and treated with intravenous heparin overnight, then discharged with anticoagulation medication. Anticoagulation was continued for at least three months after the procedure. Patients resumed the anti-arrhythmic medications they had been taking before the procedure. The patients were seen in an outpatient clinic at 1 week, and 1, 3, 6, 9, and 12 months after the procedure and then every 3–6 months thereafter.

A twelve-lead surface electrocardiogram was performed at every visit. Patients were evaluated by 24- or 48-hour Holter monitoring or 7-day event recorder at 3, 6, 9, and 12 months after the ablation and then at every six months thereafter. A detailed history of any symptoms suggesting potential AF or AT recurrence was taken. Recurrence was defined as an episode of an atrial arrhythmia of at least 30 seconds that occurred after a blanking period of 12 weeks after ablation.14

Anti-arrhythmic agents were discontinued at the three-month visit if there was no evidence of recurrence. Anticoagulation therapy was discontinued if the electrocardiogram consistently demonstrated sinus rhythm. Success was defined as the absence of any documented arrhythmia or symptoms suggestive of arrhythmia recurrence without anti-arrhythmic drugs.

Continuous variables were reported as a mean ± standard deviation, while categorical variables were reported as a number of cases and its percentage. For comparison of continuous variables, Student's t-test was used, while categorical variables were compared using χ² test or the Fisher's exact test, as appropriate. The primary end point of the study was time to first occurrence of AF. The associations between each of 16 SNPs and the recurrence of AF after catheter ablation were also examined by χ² test or the Fisher's exact test. We defined risk alleles which showed significant higher recurrence rate in selected SNP and genetic risk score was calculated after summation of the unweighted number of risk alleles of SNPs showing significant associations (P < 0.06). Cumulative AF-free survival (any recurrence of atrial arrhythmia) was estimated using the Kaplan-Meier method. Wilcoxon log rank test was used to compare AF-free survival according to the genetic risk score. To analyze independent predictors of recurrence, Cox proportional hazards regression model was employed using all variables having P < 0.05 from univariate analyses. All statistical analyses were performed using SPSS 12.0 software (SPSS Inc., Chicago, IL, USA). P values presented were two-tailed and P values < 0.05 were considered as statistically significant.

The local Institutional Review Board (IRB) of Korea University (IRB No. ED 13021) and Gachon University (IRB No. GBIRB 2015-69) approved the study, and written informed consent was obtained from all patients when they were enrolled.

A total of 89 subjects were included in this study. The baseline characteristics of the patients are shown in Table 1. The mean age at enrollment was 35.7 ± 3.7 years and the mean age at the diagnosis of AF was 33.2 ± 4.6. Eighty-one patients (91.0%) were male and fifty-seven patients (64.0%) were paroxysmal AF. Eighteen patients (20.2%) had a family history of AF and mean CHA₂DS₂-VASc score was 0.25 ± 0.46. The mean duration of AF before catheter ablation was 30.3 ± 19.2 months.

Call rates for each of the 16 SNPs were 96.6%–100%. The minor and major allele frequencies are presented in Table 2.

All patients completed at least the 12-month follow-up examination. The 12-month recurrence rate was 31.5% (28 of 89 patients) and the overall recurrence rate after single ablation procedure during long-term follow up (55.7 ± 28.3 months) was 44.9% (40 of 89 patients). Twenty-five patients (28.1%) underwent a second ablation procedure for recurred AF. Among those patients undergoing a second ablation, eight patients experienced a recurrence of AF. Three patients underwent a third ablation procedure and two of them experienced recurrence. Taken together, forty-nine patients (55.1%) were free from AF after a single procedure and the overall success rate after multiple ablation procedures was 75.3% (67 of 89 patients) after a mean follow up duration of 55.7 ± 28.3 months. Fig. 1 shows the Kaplan-Meier curve of overall AF-free survival after the final ablation procedure.

The clinical characteristics of the patients with recurred AF are compared in Table 3. The proportion of paroxysmal AF was lower in the recurrence group as compared to the non-recurrence group (38/49 [77.6%] vs. 19/40 [47.5%]; P = 0.003). The antero-posterior diameter of LA was significantly larger in patients with recurrence of AF as compared to those without recurrence (37.28 ± 6.83 mm vs. 40.02 ± 4.75 mm, P = 0.029). Procedural characteristics are summarized in Table 4. The proportion of patients requiring additional LA and RA ablation were similar according to the recurrence within each type of AF.

We examined the association between each of the 16 SNPs and recurrence rate. Wild-type rs11047543 near to SOX5 (12p12) (GG; 26/69 [37.7%] vs. GA; 13/18 [72.2%] vs. AA; 0/0 [0%], P = 0.009), rs7193343 closest to ZFHX3 (16q22) (CC; 0/7 [0%] vs. CT; 22/40 [55.0%] vs. TT; 18/41 [43.9%], P = 0.025) and the homozygous variant of rs3825214 near to TBX5 (12q24) (AA; 16/31 [51.6%] vs. AG; 22/43 [51.2%] vs. GG; 2/13 [15.4%], P = 0.056) were significantly associated with a lower rate of late recurrence while the remaining 13 SNPs were not associated with recurrence of AF (Fig. 2). Therefore, we defined the risk allele as the GA genotype of rs11047543, CT and TT genotypes of rs7193343, and AA and AG genotypes of rs3825214. When the patients were assigned to four groups according to the number of risk alleles (n = 0–3), there were significant differences in recurrence rate (n = 0; 0/3 vs. n = 1; 2/13 [15.4%] vs. n = 2; 24/52 [46.2%] vs. n = 3; 13/17 [76.5%], P = 0.003). Kaplan-Meier survival analysis showed the incremental prognostic value according to the number of risk alleles (P = 0.002) (Fig. 3).

Multivariable analysis incorporating AF type, antero-posterior diameter of LA, rs11047543, rs7139343, and rs3825214, or the number of risk alleles was performed. The non-paroxysmal type of AF was an independent predictor of recurrence (hazard ratio [HR], 2.228; 95% confidence interval [CI], 1.140–4.355; P = 0.019). The risk allele of rs11047543 (HR, 2.723; 95% CI, 1.358–5.461; P = 0.005) and the number of risk alleles (HR, 2.901; 95% CI, 1.612–5.219; P < 0.001) were significant predictors of recurrence (Tables 5 and 6).