Introduction

Many studies have confirmed that transseptal puncture leaves an unavoidable defect in the fossa ovalis. The incidence of persistent iatrogenic atrial septal defect (iASD) could depend on the size and number of sheaths inserted into the left atrium (LA), the follow-up interval, and assessment method [1,2,3,4,5,6,7]. Transthoracic echocardiography (TTE) and transesophageal echography (TEE) are ultrasound techniques used for ASD detection, and ASD color flow Doppler jet width correlates with the anatomic maximal dimension observed in the surgery [8]. TEE provides better resolution in the evaluation of atrial septum than TTE. However, TEE is invasive, requires additional sedation or anesthesia, and may be a source of some additional risk, including aspiration and esophageal trauma. Thus, TEE might not be acceptable for repeated evaluation. TTE is non-invasive, cost-effective imaging that can be performed for a considerable number of patients in a clinical examination room [9].

Recently, HotBalloon (Toray Industries, Tokyo, Japan) ablation has emerged as an alternative tool for pulmonary vein isolation (PVI). The balloon is made of elastic material, compliant enough to accommodate variations of PV anatomy, and it can be enlarged by increasing the intraballoon injection volume. In HotBalloon-based PVI (HBPVI), a 17-French balloon sheath is inserted into the left atrium (LA) after a transseptal puncture [10,11,12]. Combined use of an 8-French sheath together with the 17-French balloon sheath reduces effort in exchanging the catheter and allows us to quickly assess PV potential. Balloon ablation requires a larger sheath, and its combination with use of an electrophysiologic catheter sheath raises concern for persistent iASD. The incidence of iASD after HBPVI and related clinical symptoms after HBPVI have not yet been systematically evaluated. The present study aimed to investigate the incidence, clinical features, and predictive factors for TTE-detectable iASD after HBPVI.

Methods

Study population

This retrospective study consisted of 190 consecutive AF patients with pre- and post-ablation TTE who underwent their first PVI using a HotBalloon at our institution between June 2016 and December 2017. Post-ablation imaging was planned after 12 months to screen for iASD. If a redo ablation procedure was scheduled, post-ablation imaging was performed before the redo procedure. Patients with a previously documented atrial septal defect, any previous procedures involving puncture of the interatrial septum, or early redo procedure within 6 months after HBPVI were excluded. AF was classified according to the latest guidelines [13]. All patients gave their written informed consent, and institutional ethical approval was obtained from an ethics committee.

HotBalloon-based PVI

Deep conscious sedation was maintained throughout the procedure using propofol, thiopental, dexmedetomidine, and fentanyl. Venous access for PVI was obtained through the femoral vein. A transseptal puncture was performed using a radiofrequency needle (Baylis Medical, Montreal, Canada), targeting the fossa ovalis guided by intracardiac echocardiography. After a transseptal puncture, the puncture site was dilated using a 16 Fr dilator, and a 17-French steerable deflectable guiding sheath (Treswaltz, Toray Industries) was inserted into the LA using a guidewire. Then, an 8-French non-steerable sheath (SL0, St. Jude Medical, St. Paul, MN, USA) was inserted into the LA across a single septal puncture hole. Activated clotting time was maintained between 350 and 400 s by intermittent injection of heparin and checked every 30 min during the procedure. The ostia and branches of the PVs were identified using PV angiography during right ventricular burst pacing. Before ablation, the PV potentials in all pulmonary veins were recorded using a ring catheter (Libero, Japan Lifeline, Japan or Inter NOVA, Chiba, Japan) through the 8-French sheath. Subsequently, a HotBalloon was inserted into the LA through the 17-French sheath. The operator placed a balloon into the target PV ostium by advancing a JPC-SS 0.032″ guidewire (Lake Region Manufacturing). A diluted contrast medium (saline / contrast = 2:1) was injected to inflate the balloon through the catheter balloon lumen. Balloon positioning in the PV was adjusted through injection volume. The optimal PV occlusion by balloon wedging was verified by injecting contrast medium through the central lumen of the balloon catheter. Radiofrequency energy was delivered from the coil electrode inside the balloon by setting the central temperature of the balloon to 70 °C in the upper PV and 65 °C–70 °C in the lower PV, with an application duration of 120–240 s. Subsequently, most patients underwent carina ablation using HotBalloon. After completion of all PV ablations, the operator again recorded all PV potentials using a ring catheter. If PV potentials remained, HotBalloon ablation was repeated up to two times in each PV. If a PVI could not be completed through multiple balloon ablations, touch-up ablation for the remaining gaps was performed using a conventional radiofrequency ablation catheter.

Box isolation

In non-PAF patients, a 3D electroanatomical mapping of LA was performed. If AF persisted after HBPVI, the mapping was performed during sinus rhythm after cardioversion. A quadripolar open 3.5-mm-tip irrigated radiofrequency ablation catheter (Thermocool SmartTouch®, Biosense Webster, Diamond Bar, CA) through a 17-French steerable deflectable guiding sheath was introduced into the LA. If a low bipolar voltage area (< 0.5 mV) was present in the posterior LA wall, posterior LA wall isolation (Box isolation) was performed. For a Box lesion, a roofline was first created using a HotBalloon or a conventional radiofrequency catheter, and a floor line was then created using a conventional radiofrequency catheter alone, to prevent the risk of esophageal thermal injury. If AF still persisted after HBPVI followed by cardioversion, Box isolation was performed during AF. The endpoint of the Box isolation was when all electrical activity dissociated or was absent within the Box area, and pacing from the posterior LA wall was unable to capture the myocardium outside the Box area. Box isolation was not performed in PAF patients.

Transthoracic echocardiography

TTE (Vivid 9E, GE Medical Systems, Milwaukee, MI, USA or Philips Epic 7, Philips Medical System, The Netherlands) was performed before and after the procedure by independent sonographers. In addition to the standard measurement, an ASD was carefully evaluated in two dimensions and using the color Doppler flow from multiple views. An ASD was defined as an evident patent hole and obvious Doppler flow in an intra-atrial septum. The Valsalva maneuver was performed to transiently increase the right atrial pressure over the LA pressure. The left ventricular ejection fraction was measured using the biplane modified Simpson methods and the LA volume index was defined as the LA volume divided by the body surface area.

Follow-up

The patients were followed up at our outpatient clinic or by a general practitioner at intervals of approximately 3 months through physical examination, resting 12-lead electrocardiogram (ECG), and 24-h Holter ECG recordings. Pacemakers or implantable cardioverter defibrillators were checked if present. In cases of symptomatic palpitations, symptoms suggestive of AF recurrence, heart failure, or an embolic event, additional follow-up visits were scheduled. A symptomatic and/or documented episode of atrial arrhythmia (> 30 s) that occurred after the initial procedure was defined as recurrence. A blanking period of 3 months was applied. The subsequent need for anticoagulation depended on the CHADS2 score of the patients. [14] Discontinuation of anticoagulant therapy was considered in patients with a CHADS2 score ≤ 1 and without AF recurrence through follow-up examinations. However, the ultimate decision to discontinue anticoagulation therapy was at the discretion of physicians. If AF recurrence was detected, anticoagulation was immediately restarted.

Statistical analysis

The results are expressed as means ± standard deviations for continuous variables and frequencies and percentages for categorical variables. For comparison between the two groups, χ2 analysis or Fisher’s exact test was used for categorical variables, and an unpaired t test or Wilcoxon analysis was used for continuous variables. The atrial arrhythmia recurrence-free rates were calculated using Kaplan–Meier analysis with log-rank tests. A predictor of iASD was tested using multivariate regression analysis. Variables with p < 0.1 in univariate analyses were entered in the multivariate analysis. All tests were two sided, and p < 0.05 was considered statistically significant.

Results

Patient characteristics

The baseline characteristics of the patients are presented in Table 1. In a total of 190 patients, 98 (52%) paroxysmal AF (PAF) and 92 (48%) non-PAF patients, the mean age was 67 years and 67% of the patients were male. In non-PAF patients, the LA diameter and the LA volume index were significantly larger than those in PAF patients, and the left ventricular ejection fraction was significantly lower than that in PAF patients.

Table 1 Patient characteristics
Full size table

Procedural characteristics

The detailed procedural parameters are presented in Table 2. Overall, 131 of 190 (69%) patients were successfully isolated using HotBalloon alone, whereas the remaining 59 PVs (31%) required radiofrequency touch-up ablation. Acute electrical isolation was achieved in all PVs. Fifty-four (59%) of 92 non-PAF patients underwent Box isolation. Number of HotBalloon applications was significantly larger (8.6 ± 2.1 vs. 6.5 ± 1.4, p < 0.001) and LA procedure time was significantly longer in non-PAF patients than in PAF patients (130 ± 38 min vs. 80 ± 27 min, p < 0.001). The additional procedures of cavotricuspid isthmus ablation were performed in 23 (12%) of 190 patients.

Table 2 Procedural characteristics
Full size table

Incidence and a predictor of iASD

Follow-up TTE was performed at a mean of 12.7 ± 2.5 months after the procedure. In a total of 190 patients, 18 patients (9.4%) had a persistent iASD. A representative case is presented in Fig. 1. The size of the defect was a mean of 2.3 ± 2.0 mm. Four (22%) of 18 iASD patients were PAF, and 14 (78%) iASD patients were non-PAF. All iASD patients had left-to-right flow during the color Doppler study, and a left-to-right shunt flow disappeared during the Valsalva maneuver in one iASD patient. No patient exhibited a right-to-left shunt during the Valsalva maneuver. In the multivariate analysis, LA volume index (OR 1.044, 95% CI 1.005–1.084, p = 0.026) and LA procedure time (OR 1.025, 95% CI 1.008–1.043, p = 0.004) were identified as statistically significant independent predictors of iASD (Table 3). The receiver operating characteristic (ROC) curve analysis demonstrated that optimal cutoff value of LA volume index for predicting iASD was calculated as 54.5 mL/m2 (61.1% sensitivity and 68.2% specificity) and the area under the ROC curve was 0.684 (Fig. 2). Further follow-up TTE was performed 6 months after iASD detection in 8 of 18 iASD patients. In those 8 iASD patients, iASD was still detectable in five patients and not detectable in three patients. No patients had any symptoms related to iASD during the follow-up period.

Fig. 1
figure 1

Representative case with persistent iASD evaluated by TTE. (Left panel) short-axis view, a left-to-right shunt flow (white arrow) was detected at 10.9 months after the procedure. (Right panel) Apical four-chamber view, a left-to-right shunt flow (white arrow) was detected at 9.8 months after the procedure. iASD persistent iatrogenic atrial septal defect, TTE transthoracic echocardiograph, RA right atrium, LA left atrium

Full size image
Table 3 Predictor of iatrogenic atrial septum defect
Full size table
Fig. 2
figure 2

The receiver operating characteristic (ROC) curve of LA volume index in predicting iASD after HBPVI. The area under the ROC curve was 0.684 and optimal cutoff value of LA volume index for predicting iASD was calculated as 54.5 mL/m2 (61.1% sensitivity and 68.2% specificity). iASD persistent iatrogenic atrial septal defect, HBPVI HotBalloon-based pulmonary vein isolation, LA left atrium

Full size image

Clinical outcomes

One patient had cardiac tamponade requiring drainage during cavotricuspid isthmus ablation after HBPVI. One patient had hemoptysis during HBPVI [15]. No phrenic nerve injury, thromboembolism, or atrioesophageal fistula was observed. No intervention for iASDs was performed. During a mean follow-up of 19.1 ± 6.9 months, atrial arrhythmia-free rates in PAF and non-PAF patients 1 year after the first procedure were 76% and 61%, respectively (log-rank test, p = 0.01). The anticoagulation therapy was continued in 12 (67%) of 18 iASD patients.

Discussion

The data on iASD and related clinical symptoms after HBPVI are scarce. To the best of our knowledge, this is the first study that investigated the incidence of TTE-detectable iASD after HBPVI. The iASDs tend to spontaneously close within 3 months in the majority of cases [2]; thus, we conducted follow-up investigations 12 months after the procedures. We found that (1) 9.4% of the patients had iASD during a mean follow-up period of 12.7 ± 2.5 months, (2) LA volume index and LA procedure time were significant predictors for the incidence of persistent iASD, and (3) no patients had any symptoms related to iASD during the follow-up period. No intervention for iASDs was performed.

Incidence of persistent iASD after HBPVI

A persistent iASD is one of the complications after procedures requiring LA access. Previous studies demonstrated that the size of sheath or guiding catheter across the atrial septal was one of the strongest predictors of iASD [4]. Furthermore, a single transseptal puncture site with two sheaths is reported to be a higher risk for persistence of iASD than two separate access sites [1]. Balloon ablation requires a larger sheath, and combined use of electrophysiologic catheter sheath raises the concern of persistent iASD. In our study patients, 52% PAF and 48% non-PAF patients, a 17-French guiding sheath together with an 8-French sheath through one hole was inserted into LA. TTE-detectable iASD after HBPVI was found in 9.4% of study patients during a mean follow-up period of 12.7 months. In cryoballoon ablation, a 15-French steerable guiding sheath is inserted into LA. Cronin et al. reported that the incidence of TTE-detectable iASD at 3 months after cryoballoon ablation was higher than after a radiofrequency ablation (16.7% vs. 2.4%) [5]. In their cryoballoon population, PAF accounts for 76%, and a single transseptal puncture and a double transseptal puncture with two separate access sites were performed at similar rates. Watanabe et al. demonstrated that TTE-detectable iASD after cryoballoon ablation by a single transseptal puncture was found in 8.4% of PAF patients during a median follow-up period of 15.5 months [7]. Compared with previous cryoballoon studies using TTE, the same iASD detection method, our incidence of iASD was, not much different, at least as a reasonable result even if the differences in the follow-up period and patient characteristics were considered.

A predictive factor for persistent iASD after HBPVI

Our result indicated that the LA volume and LA procedure time were significant predictors for the incidence of persistent iASD. The relationship between LA enlargement and increased left ventricular filling pressure has been demonstrated in individuals with and without mitral valve disease; thus, LA enlargement is considered to be mostly the result of pressure overload [16, 17]. Our study population consisted of half PAF and half non-PAF patients. Importantly, 14 (78%) of 18 iASD patients were non-PAF. Non-PAF patients had large LA size, and were required longer procedure time compared with PAF patients because we performed Box isolation in 59% of non-PAF patients. Atrial septum thinning and stretching due to large LA volume, and hindrance of spontaneous closure of the atrial septum due to higher LA pressure are plausible reasons for the higher incidence of iASD in non-PAF patients. In addition, manipulation of a large sheath with longer procedure time in non-PAF patients might increase shear stress to the atrial septum hole, leading to persistence of iASD.

Future development of an intracardiac electrophysiology diagnostic catheter that can be deployed through a HotBalloon catheter can reduce the number of transseptal sheaths. Furthermore, the advancement of a smaller-diameter sheath will be also beneficial to reducing the risk of persistent iASD.

iASD-related symptoms after HBPVI

ASD could lead to systemic atherosclerotic burden, reduced cardiopulmonary exercise capacity, pulmonary hypertension, and paradoxical emboli, and manifest as dyspnea or any related symptoms, which could require closure of the ASD [18,19,20,21]. In our study, no symptoms related to iASD, embolic, or heart failure events were observed during the follow-up period. A small shunt size is a possible reason for the absence of iASD-related symptoms after HBPVI.

Our study demonstrated that non-PAF patients had higher atrial arrhythmia recurrence rates than PAF patients. Non-PAF patients accounted for the majority of iASD patients; thus, anticoagulation therapy was continued during the follow-up period in two-thirds of patients with persistent iASD. The true embolic risk for persistent iASD after HBPVI was unclear and further study will be required to determine the need for a prolonged duration of anticoagulation therapy.

Study limitations

The main limitation of this investigation is that it was a retrospective, single-center study. A transseptal puncture was targeted at fossa ovalis guided by intracardiac echocardiography. However, the exact location of the septal puncture site could not be assessed because of the retrospective nature of this study. The impact of the septal puncture site, foramen ovale, or alternative location on the persistence of iASD will require further study. The main purpose of this study was to investigate the incidence of TTE-detectable iASD, which can be measured in routine clinical practice. However, iASD might be difficult to detect by TTE, especially in small LA, and in obese and overweight patients, because of the presence of difficult echo windows. Therefore, our estimate may be a minimum value for the detection of iASDs. Long-term follow-up TTE longer than 12 months after HBPVI was not sufficient; thus, iASD that might close after 12 months could not be fully examined. Finally, all our iASD patients were asymptomatic during the follow-up period; however, long-term effects are unknown. Long-term follow-up by echocardiography is recommended for patients with iASD.

Conclusions

After HBPVI, TTE-detectable iASD was found in 9.4% of study patients. Larger LA size and longer LA procedure time were predictive factors for the persistence of iASD. All patients with iASD had no clinical symptoms 12 months after HBPVI; however, long-term follow-up may be necessary.