Association of aortic stiffness to brain natriuretic peptide in children before and after device closure of patent ductus arteriosus

Abstract

Objectives

We evaluated the influence of device closure for patent ductus arteriosus (PDA) on the aortic stiffness index (ASI) and brain natriuretic peptide (BNP) and their association with cardiac function.

Patients and methods

ASI and echocardiography assessment before and after treatment (16 ± 9 months) in 48 children with PDA (mean age 10 ± 4.5) and 52 control children (mean age 9.7 ± 4.6). BNP level was measured pre-closure for all children, and was measured six months after closure only for children with PDA.

Results

ASI was higher in PDA patients than in controls (P < 0.001). ASI correlated with age (P < 0.05), LVEF% (P < 0.01), E/E′ (<0.03), pulmonary artery pressure (P < 0.001), and BNP (P < 0.001). ASI and BNP significantly decreased after closure (P < 0.001). ASI and BNP were independent predictors for post-closure systolic dysfunction (P < 0.001and <0.005, respectively). Receiver operating curve (ROC) analysis showed that ASI ⩾ 13.5, BNP level ⩾75 pg/ml and basal mean pulmonary artery pressure (PAP) ⩾ 23 were powerful predictors for post-closure systolic function.

Conclusion

ASI is significantly associated with BNP and basal PAP in children with PDA. After device closure, aortic distensibility improved significantly and was associated with significant improvement in both systolic and diastolic functions. ASI can be used for monitoring the course of patients with PDA, and may give opportunities for early intervention.

Abbreviations

PDA

Patent ductus arteriosus

BNP

brain natriuretic peptide

ASI

aortic stiffness index

TDI

Tissue Doppler imaging

EDTA

Ethylenediaminetetraacetic Acid

AoS

Aortic systolic

AOST

Aortic strain

RAO

right anterior oblique

ADO

Amplatzer duct occluder

Introduction

Patent ductus arteriosus (PDA) causes volume overload of the left side of the heart [1] and predisposes the patient to pulmonary hypertension. The timing of treatment for congenital heart defects is based on the hemodynamic and anatomic situation, with consideration of myocardial cell adaptation and chamber remodeling. Therefore, it is important to have multiple methods available for follow-up. The combination of new imaging modalities and measurements of serum levels of natriuretic peptides may allow us to improve the evaluation of cardiac function and timing of interventions [2].

Left ventricle (LV) volume overload and compensatory remodeling alters the systolic and diastolic function of the LV as in chronic aortic and mitral regurgitation [3–5]. These changes are expected to improve after PDA closure; however, some patients develop LV systolic dysfunction. Clinical examination, X-ray chest, ECG, arterial saturation (upper and lower limbs) and echocardiography are conclusive in assessing operability in the majority of patients with PDA and pulmonary hypertension. However, the decision to intervene is difficult if the examination results are equivocal.

The purpose of this study is to evaluate the association of aortic stiffness with BNP and its relation to cardiac function before and after transcatheter closure of the PDA.

Patients and methods

Forty-eight consecutive children, who were planned for transcatheter closure of PDA, were enrolled in this prospective observational study. All patients had clinical and/or echocardiographic evidence of hemodynamically significant PDA. Patients with silent PDA, PDA not suitable for percutaneous closure, irreversible pulmonary vascular disease (pulmonary vascular resistance index, i.e., pulmonary vascular resistance index (PVRI) >7 WU m²), and those who had associated hemodynamically significant congenital heart disease or a significant residual shunt were excluded from the study. Study approval was obtained from the ethics committee and written informed consent was obtained from the parents of patients in all cases. Control subjects were examined once. They were asymptomatic and showed no abnormalities in clinical examination, ECG, or echocardiography.

Transthoracic 2D echocardiography and tissue Doppler imaging (TDI) was performed with the patient in the supine position using GE-Vivid 3 (General Electric, Milwaukee, WI, USA) with 2.5 and 3.5 MHz transducers on outpatient basis at baseline, one day after the procedure and at follow-up (at least three months after the procedure). LV systolic dysfunction was defined as a post-PDA closure absolute in left ventricular ejection fraction (LVEF) of <50% and/or reduction in LVEF of ⩾10% from the baseline. For diastolic function analysis, the mitral inflow signal was acquired from three cardiac cycles in the apical four-chamber view; the E (early mitral inflow: rapid atrial filling) and A (late mitral inflow: late atrial filling) waves were measured, and the E/A ratio was calculated. Mitral annular diastolic velocities in early diastole (E′) and late diastole (A′) were obtained by tissue Doppler imaging and E/E′ was calculated. Anatomic attributes of the ductus were assessed for size, minimum diameter toward pulmonary end, shape, orientation of PDA as well as adequacy of the ampoule on the aortic end. PDA size was measured in the parasternal short axis and ductal view. Left atrial and aortic dimensions were also measured simultaneously.

BNP measurement

All samples were collected by venipuncture into Ethylenediaminetetraacetic Acid (EDTA) tubes within two hours of obtaining the baseline echocardiogram for all children in the study and six months after device closure of PDA (for children with PDA only). The blood samples were kept at room temperature and analyzed within four hours of sampling using the Triage BNP assay (Biosite Diagnostics). In some cases, the sample was centrifuged and the plasma was frozen for one to two days at −70°C. Before analysis, each tube was inverted several times to ensure homogeneity. The BNP assay was a sandwich immuno-assay that consisted of a disposable device to which 250 mL of EDTA-anti-coagulated whole blood or plasma was added. The Triage meter was used to measure BNP concentration by detecting a fluorescent signal that reflected the amount of BNP in the sample [6]. The upper limit of the normal lab reference for BNP was 42 pg/ml [6].

Non-invasive evaluation of aortic stiffness

The transverse displacement of the aortic wall was measured with commercially available equipment (GE-Vivid 3; General Electric, Milwaukee, WI, USA), using 2.5 and 3.5 MHz transducers. After routine conventional echocardiographic examination, patients were placed in a left mild recumbent position and the ascending aorta was recorded in the two-dimensional, guided M-mode tracings. The aortic diameter was recorded by M-mode echocardiogram at a level of 3 cm above the aortic valve. Internal aortic diameters were measured by means of a caliper in systole and diastole as the distance between the trailing edge of the anterior aortic wall and the leading edge of the posterior aortic wall.

Aortic systolic (AoS) diameter was measured at the time of full opening of the aortic valve, and diastolic (AoD) diameter was measured at the peak of QRS complex of the electrocardiogram. Five consecutive beats were measured routinely and were averaged. Measurements were performed by the same investigator; the intra-observer variability of aortic diameter measurements was 2.2%. Values obtained were used to calculate aortic stiffness indices:

Aortic strain (AOST) = systolic diameter (SD) − diastolic diameter (DD)/diastolic diameter (DD)/100; Aortic distensibility = (2/aortic strain)/PP; and elastic modulus E(p) = PP/strain.

Aortic stiffness (β) Index = ln (systolic blood pressure/diastolic blood pressure)/AOST and aortic distensibility = 2 × AOST/pulse pressure (pure number) [7,8].

Cardiac catheterization

Cardiac catheterization was performed for assessment of pulmonary artery pressure and shunt quantification. Only venous access was obtained during the study. PAH was defined as mean PA pressure >25 mmHg [9]. Angiographic assessment was performed after crossing the PDA and placing a pigtail catheter in the descending thoracic aorta. Angiograms were performed in standard lateral view for PDA sizing. In selected cases, right anterior oblique (RAO) view was also used for better visualization of PDA for insertion of Amplatzer duct occluder (ADO, AGA Medical Corporation, Plymouth, MN, USA). The PDA was crossed from the pulmonary end in all patients. Amplatzer delivery sheath (AGA Medical, Plymouth, MN) was introduced from the venous route over Amplatzer super stiff guide wire (Boston Scientific, Natick, MA, USA), and was parked in the descending thoracic aorta. Device was delivered as per the standard technique described earlier [10–12]. Aortogram was done at 10 min after the release to confirm device position and rule out residual shunt. After device deployment, echocardiographic assessment was performed for the device position, descending thoracic aortic, and left pulmonary artery velocity. The ductal occluder device was released after excluding the significant residual shunt and obstruction in aorta and/or left pulmonary artery.

All cases were followed at three months and six months thereafter. Aortic stiffness and NT-BNP were re-assessed after six months. Improvement in functional class and weight gain were noted. Patients were evaluated clinically for any evidence of worsening of pulmonary hypertension. At follow-up echocardiography, the position of the device was confirmed, and residual shunt, if any, was noted.

Statistical analysis

Analyses were performed with the Statistical Package for Social Science version 12.01 for Windows (SPSS Inc., Chicago, IL). Correlation between aortic stiffness and multiple parameters was examined via Pearson and Spearman correlation coefficients.

Results

The mean age for the cohort was 9.2 ± 5.6 years at intervention, and 65.4% were females. Demographic data were represented in Table 1. They were followed for a mean duration of 23 ± 9 months after device closure. The pulmonary artery systolic pressure was less than 90% of aortic systolic pressure in all cases. All study patients had predominantly left-to-right shunt based on clinical and investigational criteria with lower limb saturations of >95%.

Table 1.

Demographic characteristics of patients with patent ductus arteriosus and controls.

Variables	Patients (n = 98)	Controls (n = 52)	P value
Age (years)	9.2 ± 5.9	9.6 ± 6.2	>0.05
Gender (m/f)	27/25	17/31	>0.05
Weight (kg)	21.8 + 18	22.7 + 19	>0.05
Hight (cm)	123.5 ± 59	119 ± 51	>0.05
Heart rate (b/m)	99 ± 12	87 ± 11	>0.05
SBP	139 ± 11.5	132.5 ± 9.3	>0.05
DBP	70.5 ± 4.5	81.2 ± 3.9	<0.05
PP	68 ± 7	50 ± 5	<0.05

SBP: systolic blood pressure, DBP: diastolic blood pressure; PP: pulse pressure.

All 48 patients with PDA underwent percutaneous closure with an Amplatzer duct occluder. In these children, no other heart defects were observed. Follow-up was available for all patients over a period of 23 ± 9 months. At the time of post deployment angiogram, 36 out of 48 patients showed residual shunt through the device. However, this disappeared over time and no residual shunt was seen in any of the patients at the time of last follow-up, where all the patients were either asymptomatic or in the New York Heart Association (NYHA) class I. During the follow-up period, none of the patients had any progression in their peak velocities or gradients across the left pulmonary artery or the aortic isthmus.

Forty-three out of 48 patients had an estimated mean PAP of ⩽20 mmHg. In the remaining five cases, though the PAP was estimated at >20 mmHg, it was significantly lower than the pre-procedural PAP. None of the patients showed any evidence of worsening PHT. At the time of the last follow-up evaluation, all the patients were either asymptomatic or in the NYHA class I. During the follow-up period, none of the patients had any progression in their peak velocities or gradients across the LPA or the aortic isthmus. BNP level was significantly higher in children with PDA than in controls (P < 0.0001) before intervention, while this decreased significantly six months after intervention (P < 0.001), becoming non-significant compared to controls (P > 0.05) (Table 2). Table 3 shows the ASI significantly correlated with age (P < 0.05), LV-EF% (P < 0.01), E/E′ (P < 0.03), baseline-PAP (P < P < 0.001) (Fig. 1), and BNP (P < 0.001) (Fig. 2). Children with PDA had significantly higher ASI than controls before occlusion (P < 0.001). After intervention, ASI decreased significantly (P < 0.003), but was still higher than in controls (P < 0.05) at the six-month follow-up evaluation (Table 4).

Table 2.

Serum level BNP in patients measured at baseline and at 6 months’ follow-up, in patients and in controls.

	Patients with PDA (N = 48)	Controls (N = 52)	P
BNP (pg/ml)-baseline	103 ± 15	32 ± 5	<0.0001
BNP (pg/ml) -6 months	39 ± 7	32 ± 5	>0.05
P	<0.001

BNP: brain natriuretic peptide.

Table 3.

Pearson’s linear correlation of aortic stiffness index with different studied parameters in patient with patent ductus arteriosus.

Parameter	Correlation coefficient (r)	P value
Age	0.229	<0.05
LV-EF%	0.436	<0.01
E/E′	0.322	<0.03
Baseline-PAP	0.615	<0.001
BNP	0.542	<0.001

LVEF: left ventricular ejection fraction; E/E′: ratio of early mitral flow velocity to early mitral annular velocity.PAP: pulmonary artery pressure; BNP: brain natriuretic peptide.

Figure 1.

Correlation of ASI and baseline pulmonary artery pressure at the time of device closure of children with PDA.

Figure 2.

Correlation of ASI and NT-proBNP at the time of device closure of PDA.

Table 4.

Two- dimensional echocardiographic findings in control subjects and in groups of children with PDA measured at baseline and at 6 and 12 months after device closuer.

	Controls (n = 52)	Patients with PDA (n = 48)		P value
		Before closure	After closure	P	P⁎
LVEDD mm	31.5 ± 5.6	36.06 + 7.4	34.9 ± 5.4	<0.001	<0.01
LVESD mm	21.9 ± 3.9	26.7 ± 5.3	22.9 ± 3.3	<0.001	<0.001
LAD mm	18.6 ± 3.0	24.7 ± 3.5	20.6 ± 3.8	<0.006	<0.03
Ao mm	19.5 ± 2.3	24.2 ± 3.1	20.5 ± 4.2	<0.02	<0.02
LA/AO	0.95 ± 0.09	1.01 ± 0.11	0.98 ± 0.09	>0.05	>0.05
LVEF%	75.5 ± 9.1	62.7 ± 4.25	70.5 ± 5.1	<0.004	<0.009
E cm/s	73.9 ± 8.3	88.3 ± 12.5	76.9 ± 9.2	<0.05	<0.05
A cm/s	61.2 ± 9.5	66.5 ± 9.8	65.2 ± 9.1	>0.05	>0.05
E/A	1.43 ± 0.3	1.41 ± 0.3	1.23 ± 0.3	>0.05	>0.05
E′ cm/s	15.8 ± 2.5	12.3 + 3.2	14.8 ± 2.5	>0.05	>0.05
E/E′	4.9 ± 1.3	7.8 ± 1.2	5.4 ± 1.3	<0.05	<0.05
Mean PAP		38.2 ± 5.3	18.5 ± 4.2	<0.001	<0.0003
AOST%	12.5 ± 2.3	8.1 ± 2.2	10.6 ± 2.0	<0.001	<0.005
Distensibility, mmHg1 .103	4.4 ± 1.5	3.1 ± 1.4	3.8 ± 0.5	<0.01	<0.01
Elastic modulus N/m2	3.9 ± 1.9	6.5 ± 2.4	4.1 ± 0.9	<0.01	<0.02
Aortic stiffness index	5.6 ± 2.5	17.5 ± 4.1	6.9 ± 1.2	<0.001	<0.001

P (Significance between patients before closure and controls).

LVEDD: left ventricular enddiastolic diameter; LVESD: left ventricular endsystolic diameter; LVEF: left ventricular rjection fraction; E/E′: ratio of early mitral flow velocity to early mitral annular velocity; PAP: pulmonary artery pressure; AOST: aortic strain.

^⁎P (Significance between patients before and after closure).

The mean pre-closure LVEF% of the children with PDA was 62.7 + 4.25%. The mean 48 h post-closure LVEF% was 55.2 ± 5.4%; while the mean LVEF% after six months follow-up was 70.5 + 5.1. Of the 48 patients, 12 (25%) had a reduced LVEF% post-closure (EF% <50% or decrease >10% below the pre-closure EF%). Analysis of patients with reduced EF% showed they had higher ASI (P < 0.001), BNP levels (P < 0.001), higher pre-closure mean PAP (P < 0.003), and duct size (P < 0.005) as compared to those with normal systolic function (Table 5). The mean time to normalization of systolic function was determined at 6 ± 1.8 months. In one patient only, the device embolized in the descending aorta, and was successfully snared and re-implanted.

Table 5.

Comparison between children regarding post-closure left ventricular systolic function.

Preclosure variable	Children with immediate systolic dysfunction (n = 12)	Children with normal systolic function (n = 36)	P value
Aortic beta index (β)	21.5 ± 2.9	13.6 ± 2.8	<0.001
BNP	127 + 23	81 + 14	<0.001
PDA size mm	9.8 ± 1.9	5.4 ± 1.2	<0.005
LVEF%	45% + 3	65% + 4	<0.002
Mean PAP mmHg	32 ± 8	17.2 ± 5	<0.003
E/E′	9.7 ± 2.4	6.2 ± 1.8	<0.02

BNP: brain natriuretic peptide; PDA: patent ductus arteriosus’; LVEF: Left ventricular rjection fraction; PAP: pulmonary artery pressure; E/E′: ratio of early mitral flow velocity to early mitral annular velocity.

The mean duct size, as determined by either echocardiography or during intervention, was 8.6 ± 4.1 mm. The mean PDA size was larger (9.8 ± 1.9 mm) for children who had higher ASI and elevated levels of NBP as compared to children with smaller PDA size (5.4 ± 1.2 mm). All patients were NYHA class II.

In order to identify the factors associated with a post-closure decrease in the EF% <50% or more than 10% below the pre-closure EF%, stepwise multiple linear regression analysis was performed (Table 6). Univariate stepwise linear regression analysis showed that age, EF%, E/E′, baseline PAP, ASI and BNP level were significantly associated with LV systolic dysfunction at post-closure examination. Among these parameters, the ASI and BNP were statistically significant factors on multivariate stepwise linear regression analysis for post-closure LV systolic dysfunction (P < 0.001and <0.005 respectively).

Table 6.

Multiple linear regression analysis for left ventricular dysfunction at post-closure.

	Univariate		Multivariate
Pre-closure variables	r	P value	r	P value
Age (years)	0.346	<0.05	0.115
LVEF%	0.464	<0.02	0.137
E/E′	0.397	<0.03	0.192
Baseline-PAP (mmHg)	0.513	<0.001	0.210
ASI (B)	0.562	<0.001	0.516	<0.001
BNP (Pg/ml)	0.541	<0.001	0.509	<0.005

LVEF: Left ventricular rjection fraction; E/E′: ratio of early mitral flow velocity to early mitral annular velocity; PAP: pulmonary artery pressure; ASI: aortic stiffness index; BNP: brain natriuretic peptide.

By ROC analysis (Table 7), the area under the curve for ASI was 0.95, and pre-closure ASI (⩾13.5) showed a sensitivity of 86% and specificity of 91% in predicting the post-closure normal LV systolic dysfunction. Pre-closure BNP ⩾75 pg/ml showed a sensitivity of 82% and specificity of 85%; mean PAP ⩾33 mmHg showed a sensitivity of 79% and specificity of 82%; and pre-closure LVEF ⩽55% showed a sensitivity of 85% and specificity of 89% in predicting the immediate post-closure normal LV systolic function.

Table 7.

ROC curve of variables for postclosure LVEF.

	Area under ROC	Sensitivity (%)	Specificity (%)
Aortic beta index ⩾	0.95	86	91
BNP ⩾	0.93	84	89
Mean PAP ⩾33 mmHg	0.811	79	82
LVEF ⩽ 55%	0.92	83	87

BNP: brain natriuretic peptide; pulmonary artery pressure; LVEF: left ventricular ejection fraction.

Discussion

The main findings in the current study are: (1) Aortic stiffness was significantly increased in children with PDA; (2) ASI was positively correlated with pre-closure LVEF%, BNP, E/E′ ratio, basal PAP and age; and (3) aortic stiffness decreased significantly after device closure of PDA, with associated significant decrease in the level of BNP with improvement of both systolic and diastolic functions.

The timing of treatment for congenital heart defects is based on the hemodynamic and anatomic situation, with consideration of myocardial cell adaptation and chamber remodeling. Therefore, it is important to have multiple methods available for follow-up. We hypothesized that the biophysical properties of the aorta are abnormal in patients with PDA, and increased arterial stiffness is an additional factor responsible for the development of late ventricular failure in those patients.

Given that evidence of patent ductus arteriosus is usually associated with hyperdynamic status and that there is a vascular shunt between the aorta and pulmonary artery, intrinsic aortic changes occur (aortic stiffness). Moreover, shunt lesions may be associated with an inflammatory process, and endothelial dysfunction may accelerate the aging of vessels, especially the aorta (aortic stiffness) [13].

Hemodynamic changes, oxygen saturation changes (nocturnal hypoxemia) due to rise of pulmonary artery pressure, in addition to the inflammatory profile and endothelial dysfunction, may be the underlying mechanisms of aortic stiffness in shunt lesions [14,15].

Outcome after closure of a large PDA is determined primarily by age at the time of repair and pre-operative pulmonary vascular disease [16]. Age is an important predictor of pulmonary vascular disease. It is generally agreed that children under one year of age are unlikely to have irreversible PAH, and most agree that irreversibility starts as early as one to two years of age. This generalization has some limitations as the pathogenesis of irreversible PAH and its progression is multifactorial and unpredictable. Blount and associates [17]. indicated that a PDA may have a more marked effect on the pulmonary circulation than on a ventricular septal defect, and that irreversible pulmonary vascular changes may occur under two years of age. This is probably the result of the high-pressure pulsatile flow transmitted from the aorta to the PA throughout the cardiac cycle in PDA.

At the level of the ascending aorta the results may differ and aortic stiffness in patients with PDA could be higher than in matched control subjects, considering the possible damage to the endothelium due to turbulent flow transmitted from the PDA. This could also explain the improvement of aortic stiffness after device closure of PDA. This improvement in aortic stiffness could contribute to the reverse remodeling of left ventricle.

BNP levels were higher in patients with PDA than in controls at baseline. The peptide level was elevated 4.3 ± 2 months after treatment, and decreased thereafter. At the six-month follow-up evaluation, the dimensions of the originally overloaded ventricle in patients with PDA had normalized.

Studies show that a day after PDA occlusion, serum levels of NT-proBNP increase, but decrease to control levels in six months [18]. The highest levels of NT-proBNP in patients with PDA and volume overload of LV are in accordance with the previous studies [18,19]. To our knowledge, there are no published reports of a correlation between ASI and peptide levels in patients with left-sided volume overload caused by PDA. In our patients with PDA, a significant correlation was found between serum levels of BNP and the aortic stiffness index. Plasma levels of BNP have been demonstrated to correlate with systolic right ventricle (RV) pressure in children with volume overload of RV [20]. Plasma levels of BNP have been shown to correlate with right atrial and ventricular pressures in a child population consisting of different loading conditions and a wide age range [21].

Once closed, hemodynamically significant PDA decreases the preload to the LV by abolishing the left-to-right shunt, and increases the afterload by isolating the low-resistance pulmonary circulation from LV outflow circulation. This simultaneous reduction in the LV preload and increase in the afterload may lead to LV systolic dysfunction. Despite similar hemodynamic changes in patients with PDA, LV systolic dysfunction is not a universal feature. Similar LV systolic dysfunction has been documented in the postoperative period of mitral regurgitation surgeries [3,4]. The incidence and predictors of post-PDA closure and LV systolic dysfunction have been reported in recent studies [22–24]. Prior studies have shown that subjects with PDA have higher LVESVI and LVEDVI, lower LVEF, and higher N-terminal brain natriuretic peptide (BNP) compared to controls. These changes are documented as resolved over a six-month follow-up period after percutaneous PDA closure [18].

It is important to note that a small percentage of patients with borderline hemodynamic data with PDA and PAH can deteriorate after PDA closure due to non regression of pulmonary hypertension and progressive PVD and right heart failure after PDA closure, and their natural history is then similar to primary or idiopathic PAH. These patients have a more favorable natural history if PDA is left untreated. A foolproof investigation to identify who may benefit from PDA closure with long-term regression of PAH, and who may worsen with progressive pulmonary vascular disease and right heart failure is currently not available. Future research on the type and extent of morphological changes in the pulmonary vessels, individual variability and correlation with genetic and epigenetic factors may give a clue to this vexing issue. Until such time, in clinical practice, an occasional patient who will not benefit from closure of a large PDA may have PDA closure with an adverse outcome and others without an adverse outcome

Limitation

The most significant limitation to this study is that the number of patients is small. However, the age range among patient groups and controls is wide.

Clinical relevance

We have demonstrated that LVEF%, baseline sPAP and E/E′ ratio remain in correlation with noninvasive measurement of arterial stiffness in patients with PDA. The association of elevated BNP and increased arterial stiffness and therefore elevated cardiovascular risk are important in the assessment of early intervention for patients with congenital shunt lesions such as PDA. It has been postulated that screening tests that could identify earlier stages of cardiac dysfunction development would be of great value. The combination of noninvasive measurements of arterial stiffness and BNP level as predictors of systolic dysfunction could be helpful in the comprehensive care of patients.

Conclusion

Aortic stiffness is significantly elevated and associated with higher BNP levels in patients with PDA. After device closure, ASI decreased significantly and was associated with significant decrease in BNP levels. These changes were associated with significantly improved systolic function months after device closure. ASI might be useful for monitoring the course of patients with PDA before and after intervention.