Effect of Renal and Left Ventricular Function on Serial Pulmonary Arterial Pressure Changes after Device Closure of Atrial Septal Defect

Background The age of candidates for device closure of atrial septal defect (ASD) has been increasing. Thus, concerns exist about dyspnea aggravation or atrial fibrillation development after device closure due to augmentation of left ventricular (LV) and left atrial (LA) preload. This study aimed to examine patterns and determinants of serial pulmonary arterial pressure and left ventricular filling pressure changes after device closure of ASD. Methods Among the 86 consecutive patients who underwent percutaneous device closure of ASD, those with end-stage renal disease or those without pre- or postprocedural Doppler data were excluded. The clinical, transesophageal, and transthoracic echocardiographic findings of 78 patients were collected at baseline, one-day postprocedure, and one-year follow-up. Results The mean age of study patients was 49.8 ± 15.0 years, and the average maximal defect diameter and device size were 20.2 ± 6.0 mm and 23.8 ± 6.4 mm. Four patients (5.6%) underwent new-onset atrial fibrillation, and five patients (6.4%) took diuretics within one-year after closure. Some patients (n = 21; 27%) exhibited paradoxically increased tricuspid regurgitant velocity (TRV) one-day postprocedure; they also were older with lower e', glomerular filtration rate, and LV ejection fraction and a higher LA volume index. However, even in these patients, TRV deceased below baseline levels one-year later. Both E/e' and LA volume index significantly increased immediately after device closure, but all decreased one-year later. Larger defect size and higher TRV were significantly correlated with immediate E/e' elevation. Conclusion In older, renal, diastolic, and systolic dysfunctional patients with larger LA and scheduled for larger device implantation, peri-interventional preload reduction therapy would be beneficial.


Introduction
e age of candidates for device closure of atrial septal defect (ASD) has been increasing. Especially in older patients, concerns exist about dyspnea aggravation [1] or atrial fibrillation development after device closure due to the augmentation of left ventricular (LV) and left atrial (LA) preload [2,3]. Although recent studies showed that the long-term outcome of device closure is quite good even in older patients, vulnerable patients must be identified to prepare them for periprocedural preload reduction management. us, in this study, we sought to examine serial hemodynamic changes, including pulmonary arterial systolic pressure, LV filling pressure, stroke volume, and LA volume, after device closure of ASD. In addition, factors contributing to adverse hemodynamic changes from immediately after the procedure to one-year follow-up were determined.

Study Population and Clinical Follow-Up.
e database of a single center of patients who underwent percutaneous device closure of ASD was analyzed. Patients underwent transthoracic echocardiography (TTE) and transesophageal echocardiography (TEE) before the device closure procedure. Among the 86 consecutive patients who underwent percutaneous device closure of ASD, patients with end-stage renal disease or patients without pre-or postprocedural Doppler data were excluded. In total, the clinical, transesophageal, and transthoracic echocardiographic findings of 78 patients were collected at baseline, one day (immediately) after the procedure, and one-year follow-up ( Figure 1). All patients underwent blood chemistry and cell blood count analysis. Estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration creatinine equation [4]. Clinical characteristics, medical history, laboratory findings, medications, and echocardiographic data were collected and used for analysis. e study protocol was approved by the institutional review board of Gangnam Severance hospital (3-2020-0026), and the need for written informed consent was waived due to the study's retrospective design.

Echocardiography at Baseline and after Device Closure.
Each patient underwent a complete standard TTE. TTE findings were collected at baseline, one-day (immediately) postclosure, and one-year follow-up. LV dimensions and septal and posterior wall thickness were measured at enddiastole and end-systole in the two-dimensional (2D) parasternal long-or short-axis views. LV ejection fraction was calculated using the modified Quinones' method [5]. LV mass was measured by Devereux's methods as recommended by the American Society of Echocardiography [5]. LA volume was measured using the prolate ellipsoidal method at the point of LV end-systole at maximum LA size. From the apical window, a 1 mm pulsed Doppler sample volume was placed at the mitral valve tip, and mitral flow velocities from 5-10 cardiac cycles were recorded. Peak early (E) and late (A) mitral inflow velocities were also measured. Mitral annular velocity was measured by tissue Doppler imaging using the pulsed-wave Doppler mode. e filter was set to exclude high-frequency signals, and the Nyquist limit was adjusted to a range of 15-20 cm/s. Gain and sample volume were minimized to allow for a clear tissue signal with minimal background noise. Systolic (S') and early (e') and late diastolic velocities of the mitral annulus were measured from the apical four-chamber view with a sample volume (2-5 mm) placed at the septal corner of the mitral annulus. Peak velocity of tricuspid regurgitation was measured. Pulmonary arterial systolic pressure (PASP) was calculated as follows: 4 × tricuspid regurgitant velocity (TRV) 2 + right atrial pressure, where right atrial pressure was estimated according to inferior vena cava diameter and its respiratory variations [6]. E/e' divided by stroke volume was defined as LV end-diastolic elastance index [7]. During TEE, multiplane 2D and zoomed 3D images were acquired, and both long-and short-axis diameters of the ASD were measured. Area was calculated as long-axis diameter multiplied by short-axis diameter and 3.14. In multiple defects, the sum of all diameters or areas was used. e margins to the aorta, posterior wall, superior vena cava, and inferior vena cava rims were also evaluated.

Device Implantation and Periprocedural Imaging.
Transcatheter ASD closure was performed, as described previously [8], using various types of septal occluders (Amplatzer/Cocoon/Figullar Flex II/Gore-Helix/Occlutech). Before the procedure, the pulmonary-to-systemic blood flow ratio and pulmonary artery pressure were evaluated using cardiac catheterization. During the procedure, 3D-TEE or intracardiac echocardiography (ICE) were performed to guide accurate device implantation and ensure successful device closure. After the procedure, all patients received 100 mg/day aspirin for at least 6 months and 75 mg clopidogrel for 3 months. Other medications, such as diuretics and antihypertensive medications, were continued.

Statistical Analysis.
Clinical characteristics and echocardiographic parameters are presented as the means ± standard deviation for continuous variables and the numbers (percentage) for categorical variables. Correlation analysis was performed between continuous variables using the Pearson correlation coefficient. Comparisons of clinical and echocardiographic findings between two groups were performed using the independent t-test. Nonparametric comparisons of two groups were done by Mann-Whitney U test. Serial changes in echocardiographic parameters were assessed using repeated ANOVA or the paired t-test. Variables with P values less than 0.05 in univariate analysis were included in the multivariable linear or logistic regression analysis. All the analyses were performed using SPSS (version 25.0, IBM, USA), and P values less than 0.05 were considered significant.

Baseline Characteristics, Echocardiographic Parameters, and Hemodynamic Findings.
e mean age of enrolled patients was 49.8 ± 15.0 (range, 16-77) years, and 51 (65%) were female. e average body mass index was 22.5 ± 3.0 kg/ m 2 , and the average eGFR was 99.4 ± 20.0 mL/min. Mean LV ejection fraction and LV mass index were 66.1 ± 6.4% and 63.3 ± 16.3 g/m 2 , respectively. Among them, 4 patients had more than one defects. Measured average defect maximal diameter was 20.2 ± 6.0 mm, and defect area was 2.57 ± 1.52 cm 2 on preprocedural TEE. e calculated Qp/Qs ratio during right side catheterization before device implantation was 2.52 ± 0.85. Sixty-four patients underwent TEE, and 14 patients underwent ICE for periprocedural guidance.
e average size of the septal occluder was 23.8 ± 6.4 mm. Baseline characteristics are described in Table 1.

Clinical Problems after Device Closure.
Six patients had history of persistent atrial fibrillation or documented paroxysmal atrial fibrillation. Among the rest (n � 72), four patients (4/72, 5.6%) underwent newly developed or detected atrial fibrillation within one year after device closure. Although it is not statistically significant due to small number of the patients, the patients with new-onset atrial fibrillation after closure had tendency of older age, lower eGFR, e', and LV ejection fraction, and larger defect size, LA volume index, and E/e. We could not find a case with development of significant pulmonary edema; however, five patients underwent taking diuretics after device closure. ey had higher prevalence of atrial fibrillation, E/e', and lower eGFR, tendency of older age, lower eGFR, e', and higher LA volume index and E/e' without significance (Table 2).

Serial Changes in LV Filling
Pressure, LA Volume, and Stroke Volume. At baseline, E/e' and LA volume index were 9.0 ± 3.5 and 30.5 ± 12.5 mL/m 2 , respectively. One day after the procedure, these values significantly increased to 11.4 ± 4.0 and 32.5 ± 12.1 mL/m 2 , respectively, due to increased LV filling from occlusion of a left-to-right shunt. LV stroke volume increased from 52.1 ± 11.2 mL to 61.5 ± 19.4 mL the day after the procedure. At one-year  follow-up, E/e' and LA volume index decreased by 9.8 ± 3.0 and 27.8 ± 8.6 mL/m 2 , which was accompanied by a further increase in LV stroke volume (Figure 2). Immediate E/e' changes after device closure (ΔE/e'-immediate) were significantly correlated with immediate LA volume index change (r � 0.424; P < 0.001). ASD area, maximal defect diameter, and implanted device diameter were significantly correlated with ΔE/e'-immediate but not with age. Higher TRV and lower baseline E/e' were correlated with ΔE/e'immediate (Table 3). In multivariate analysis, TRV and baseline E/e' were related to ΔE/e'-immediate.

Serial Changes in Pulmonary Arterial Pressure after Device
Closure. Of the 78 patients enrolled in the study, in two patients, pre-or post-TRV were not measurable and 39 underwent TTE at one-year follow-up. e average PASP decreased approximately −6.15 ± 11.6 mmHg immediately after closure, but some patients (n � 21; 28%) paradoxically exhibited increase in PASP. Patients with immediately increased PASP were older, with lower eGFR, LV ejection fraction, and e' and higher LA volume index (Table 4). In multivariate analysis for increased PASP-immediate, LV ejection fraction was significantly related to PASP increase immediately after closure. When serially followed, these patients also showed serially decreased PASP one-year later below the baseline level, similar to patients with initially decreased PASP (Figure 3). In patients with immediately increased PASP, accompanying increase in LA size was significantly blunted compared to decreased PASP group (Figure 4).

Discussion
According to our study results, we found some patients underwent new-onset atrial fibrillation and took diuretics within one-year after device closure. In addition, some patients who underwent device closure of ASD exhibited immediate PASP elevation even after transpulmonary arterial flow was reduced by blocking left-to-right shunt flow.
is finding may be from immediately elevated pulmonary capillary wedge pressure due to increased LV filling flow to the noncompliant left ventricle or volume overload to the noncompliant left atrium (perhaps from not only the muscular part but also the device-covered noncompliant part) [9]. is speculation was supported by our findings that impaired LV relaxation (as represented by lower e') and systolic function, and higher LA volume index was found in the initial increased PASP group.
is finding was more predominant in older patients with lower eGFR, suggesting that these patients need peri-interventional preload reducing medication, especially those with impaired diastolic function and larger LA size. When elevated LV filling pressure measured by E/e' was seen after device closure, larger defect size was more related to LV filling pressure elevation. us, especially in patients scheduled to close a larger defect, physicians should be cautious to avoid postinterventional pulmonary edema or exertional dyspnea aggravation. When considering the poor correlation between delta PASP elevation and delta E/e' elevation after device closure, increased PASP would be contributed by not only elevated LV filling pressure [10] but also LA noncompliance, which is also contributed by the device itself in the LA septum. An immediate increase in E/e' was significantly correlated with increase in LA volume index. However, at one-year followup, decreased LV filling pressure was accompanied by decreased LA volume index. When examining serial changes at one-year follow-up, we found that all patients showed decreased PASP below the baseline level regardless of the immediate PASP response, even in the paradoxically increased PASP group. us, despite the concern of device closure in older patients with renal dysfunction, these findings support that long-term hemodynamic response   would be favorable in older patients. erefore, with optimal preload reduction and volume control during and immediately after the procedure, device closure of ASD can be recommended even in vulnerable patients [11]. is study has some limitations. Although we found a significant proportion of patients underwent paradoxically increase in PASP, especially in older patients, with renal dysfunction, impaired LV relaxation, and larger LA, the contribution of the device on the LA side could not be fully determined. We could only assume that, with larger LA and marginal compliance, adding a prosthesis may worsen LA compliance. Although we could not find any difference in device type, future studies are warranted to determine how to reduce worsening LA compliance when selecting device size, device type, and deployment method.

Conclusions
In older patients with impaired LV relaxation, systolic dysfunction, renal dysfunction, or larger LA size, PASP could paradoxically increase after device closure of ASD due to the immediate volume overload to a noncompliant left ventricle and left atrium. erefore, periprocedural preload manipulation, such as diuretics, is recommended in patients with these risk factors who are scheduled to close a larger defect. However, after the immediate periprocedural period, TRV, LA volume index, and E/e' continuously decrease, suggesting the favorable effects of ASD closure even in these high-risk patients. e risk for significant symptoms of heart failure is small with due precautions as none of the patients in this cohort experienced any.