Transcatheter device closure of patent ductus arteriosus without arterial access – Single institution experience

Abstract

Objective

Retrospective analysis of feasibility, safety and advantages of device closure of patent ductus arteriosus (PDA) using only venous access.

Background

Arterial access for transcatheter device closure of PDA has been a standard practice, but has inherent complications, especially in infants.

Method

Records of patients who underwent PDA device closure from 2004 to 2012 were reviewed. Echocardiography was used for patient selection and for assessment of procedural outcome.

Result

151 out of 179 patients underwent PDA device closure with venous access alone, weighing 2.2–58 kg with half <10 kg and follow up of 6 months–8 years. Fluoroscopic time ranged from 2.2 to 16 min. Immediate closure was achieved in 146 patients. Two patients had new-onset left pulmonary artery turbulence and one had residual flow.

Conclusion

PDA device closure without arterial access can be accomplished safely and effectively in vast majority of patients including infants.

1. Introduction

Patent ductus arteriosus (PDA) that is moderate or large requires attention in infancy because of symptoms of heart failure, frequent respiratory infections and failure to thrive.^1,2 Arterial access is often obtained during PDA device closure to profile the duct and to assess the outcome of device deployment. This requires heparinization, as it carries a small risk of femoral arterial occlusion. Since the introduction of systemic heparinization in 1970, the incidence of femoral artery thrombosis has decreased but incidence of femoral artery related complications cannot be ignored, especially in children less than 10 kg weight.^3,4 Apart from thrombosis, other potential issues include bleeding, longer catheterization and fluoroscopic time and probably prolonged hospital stay. We present a comprehensive review of transcatheter PDA device closure in 179 patients over a period of eight years, out of which 151 (84%) underwent successful closure of PDA without femoral arterial access.

2. Methods

2.1. Study design

From August 2004 onwards, we have been attempting to close PDA using venous access alone. The records of all patients who underwent transcatheter closure of PDA between 1st August 2004 and 31st May 2012 at our institution were reviewed. Patients in whom the intention was to close the duct without arterial access were the focus of this study.

2.2. Patient selection

Device closure was attempted in patients with hemodynamically significant PDA. A detailed echocardiographic evaluation of the duct morphology was performed for all patients undergoing PDA closure as part of an institutional protocol. This was performed using broadband 1–3 MHz, 3–8 MHz and 5–12 MHz frequency transducers (Sonos 7500, Philips). The duct diameter was measured in a high parasternal long axis view at the point where it opened into the pulmonary artery. Since this measurement varies with different phases of cardiac cycle, the maximum measured diameter at pulmonary artery insertion was reported. Repeated measurements were made and the most concordant measurement was taken. The ductal ampulla was defined in the high parasternal long axis or in the suprasternal long axis view. The ductal ampulla was considered adequate if its maximal dimension along the long axis was greater than twice the measured ductal diameter. Careful two-dimensional and Doppler imaging of the pulmonary artery branch origins was also routinely performed. After the echocardiogram, the device closure was planned. The device size was selected as 2 mm more than the measured duct size.

Device closure using a venous access alone was planned for all patients except the following situations when arterial access was considered necessary and was obtained on an elective basis:

• –Patients in whom echocardiographic definition was considered inadequate.
• –Patients with large ducts with severe pulmonary artery hypertension which needed a diagnostic study to assess pulmonary vascular resistance (PVR) for reversibility prior to attempted device closure.
• –Small infants with large ducts requiring bigger devices which can potentially cause descending aortic obstruction. In this subset of patients, the strategy which we used was to measure the transverse descending aortic diameter just below the duct and if this was less than or equal to the aortic retention disc (device preselected by duct size on echocardiographic assessment), the arterial access was obtained.
• –Patients with additional lesions requiring intervention.

Informed consent was obtained. All procedures were performed under conscious sedation with intravenous ketamine and midazolam and local anesthesia. None of the patients received general anesthesia. The blood pressure was monitored noninvasively. Heparin was not administered at the start of the procedure.

2.3. Angiographic definition

A 5 French multipurpose catheter was used to cross the duct from the main pulmonary artery with the help of the straight end of a 0.025 or 0.035 inch guidewire (Terumo Medical Corporation, Somerset, NJ). The catheter was exchanged with an appropriate sized long sheath (Cook Medical Inc., Bloomington, IN) (according to the preselected device size on echocardiography) over an exchange length extra-stiff guidewire. The duct was profiled in lateral (Fig. 1) and if required, in right anterior oblique view (Fig. 2) by hand injection with the tip of the long sheath positioned within the ductal ampulla or the proximal descending aorta with the stiff wire still across the sheath. The retrograde blood flow into the pulmonary artery from the ductal ampulla and proximal descending thoracic aorta allowed definition of the pulmonary artery insertion of the duct. The ductal size was measured and was compared with our echocardiographic measurements.

Fig. 1.

Hand injection through long sheath in lateral projection delineating PDA via retrograde filling; a conical shaped structure with wide end towards aorta and tapering end (white arrow) towards pulmonary artery. MPA – main pulmonary artery.

Fig. 2.

Hand injection through long sheath in right anterior oblique view showing anatomy of PDA with narrow end (white arrow) towards pulmonary artery. MPA – main pulmonary artery.

2.4. Device selection

Device size was selected based on echocardiographic measurements if both echocardiographic and angiographic values were concordant, otherwise the bigger measurement was taken. The devices used for ductal closure were Amplatzer Duct Occluders (AGA Medical Corporation, Golden Valley, MN), Cardi-O-Fix Duct Occluders (Starway Medical Corporation) and Lifetech Duct Occluders (ShenZhen Lifetech Scientific Inc.). The devices were selected according to their availability and cost considerations.

2.5. Device deployment technique

After angiography, the long sheath was repositioned in the descending aorta and the device was deployed according to the standard technique using tracheal air shadow as the landmark for duct ampulla and opening of the duct into the pulmonary artery. The PA pressure was measured after deployment of the device through the side arm of the long sheath. The origins of the branch pulmonary arteries were assessed by hand injections through the long sheath in anteroposterior and in left anterior oblique view with slight cranial angulation (Fig. 3). In selected patients, this was repeated even after release of the device, especially in smaller children.

Fig. 3.

Hand injection through long sheath in left anterior oblique view with cranial angulation after placement of device (delivery cable still attached) showing normal filling of branch pulmonary arteries. MPA – main pulmonary artery; LPA – left pulmonary artery; RPA – right pulmonary artery.

The special techniques we have used were:

• •To use the Amplatzer perimembranous ventricular septal defect occluder delivery cable (AGA Medical Corporation, Golden Valley, MN) for deployment of devices in small children. We found this to be secure and easy to track the device through the curved portion of the long sheath. The cable was compatible with Amplatzer and Lifetech duct occluders.
• •In case of kinking of the sheath, we position the sheath deep in the descending aorta and then track the device to the tip of the sheath before deploying it in the appropriate position.

Before releasing the device, echocardiography was performed in the cardiac catheterization laboratory. The device placement was assessed on 2D echocardiography in high parasternal view and suprasternal view. The correct placement of the device was accepted when the retention skirt was seen opposite the ampulla in the descending aorta and the conical portion of the device was seen in the pulmonary artery with a waist at the mouth of the duct (Fig. 4). Color flow mapping was done to check for residual flow through the duct as well as turbulence at the origin of the left pulmonary artery and in the descending aorta (Fig. 5). Small residual color flow was accepted. Doppler gradients of upto 5 mmHg in the left pulmonary artery and upto 10 mmHg in the descending aorta were accepted. The device was deployed and the echocardiographic assessments were repeated.

Fig. 4.

Echocardiographic assessment of device placement in high parasternal view. The retention skirt seen occupying the ampullary space and the conical part projects into pulmonary artery across the duct. RPA – right pulmonary artery.

Fig. 5.

Colour Doppler interrogation in (a) parasternal short axis view showing no residual flow across the duct (b) parasternal short axis view showing normal LPA flow (c) suprasternal long axis view showing unobstructed flow in descending aorta. LPA – left pulmonary artery; RPA – right pulmonary artery.

All patients underwent repeat echocardiography before discharge, where all the above mentioned parameters were reassessed.

2.6. Discharge and follow up

Patients were discharged 24 h after the procedure if it was acceptable to the parents. All patients were asked to return for a follow-up echocardiogram 1 month after the procedure for assessment of residual flow across the duct and to check for turbulence at the origin of the left pulmonary artery and descending aorta. They were then assessed at regular interval of six months till 3 years after the procedure. Complete evaluation including echocardiography was done at each visit.

3. Results

From August 2004 to May 2012, we attempted PDA device closure in 179 patients. Device closure using venous access alone was achieved in 151 patients [age: 2 months – 34 years (median 2.7 years) with 50 patients <1 year of age; weight: 2.2–58 kg (median 12.4 kg), including 85 patients with weight < 10 kg]. The PDA sizes ranged from 1.8 mm to 14 mm (mean: 4.5 ± 2.3 mm).

Arterial access was required in 28 patients and the reasons were as follows:

• 1.Elective arterial access (19 patients), for reasons explained before
• 2.Accidental arterial puncture while trying to obtain venous access (4 patients)
• 3.Unsatisfactory echocardiographic images for assessment of results after deployment of the device (3 patients)
• 4.Unable to obtain venous access (1 patient)
• 5.Unable to cross into PDA from PA (1 patient).

The flow across the PDA could be completely eliminated in the catheterization laboratory in 146 of the 151 patients. The residual flow in 4 patients disappeared 24 h after the procedure. Turbulence at the LPA origin was demonstrable in 8 patients out of 151. Three of them had mild LPA origin stenosis demonstrated prior to PDA closure which remained same after the procedure. Five patients had new-onset LPA turbulence (peak gradient range 3–15 mmHg). In two patients, the gradients appeared after device release (there was no turbulence while the cable was attached). Flow at the LPA origin remained laminar in the remaining patients. Two patients had new-onset turbulence in the descending aorta with peak gradients of 8 mmHg and 10 mmHg.

The mean fluoroscopic time for the procedure was 3.5 ± 4.3 min (range: 2.2–16 min). No other complications or procedural events were recorded. Of patients who required arterial access, two patients needed overnight heparin for impairment of lower limb perfusion. So, the need for heparinization for vascular compromise for the whole group was low (2/179) – 1.1%.

Out of 151 patients, one-month follow up data was available for 147 patients. Residual PDA flow was seen in one patient. The gradient in LPA remained unchanged in the 5 patients including 3 patients who had LPA stenosis prior to the procedure. The aorta gradients disappeared on follow-up at one year in both patients.

4. Discussion

The first transcatheter closure of PDA was reported by Porstmann et al in 1971.⁵ Closure of PDA is now a well established and effective procedure for which various devices have been used.^6–10 Interventional closure of PDA is considered as safe method of duct occlusion but some complications such as hemolysis, device embolization, infection, and significant narrowing of the left pulmonary artery or the descending aorta have been reported in the past.^6,11,12 Device embolization has been identified as one of the most significant complications of interventional PDA occlusion.^13,14 The reported embolization rate varies, between 0% and 16%.^10,15,16 The reported rate of hemolysis varies between 0% and 3.5%.⁶

Routine practice involves arterial access for angiography before, during, and after deployment of the device. The reasons for using arterial access is that it allows initial aortography to be performed without crossing the duct, supposedly minimizing the risk of ductal spasm and consequent underestimation of the size of the duct. Also, it permits subsequent aortography to confirm a good position of the device prior to release and to exclude iatrogenic coarctation. It is our experience, and that of others,¹⁷ that ductal spasm may occur whether the duct is crossed or not. Arterial access in children is reported to have a high rate of complication ranging between 3.7 and 16%, with a significant proportion of complications requiring intervention.^3,18 Complications such as arterial disruption, or acute occlusion, may be limb- threatening.¹⁹ Although the use of thrombolytics has been found to be effective in treating femoral arterial thrombosis,¹⁹ a proportion of children still have absent or diminished pulses in the foot at the time of discharge, and the drugs themselves are not risk-free. Chronic occlusion of the ilio-femoral arteries after catheterization can result in claudication, inequality in the length of the legs, and disturbance of gait that is particularly important in children with growing epiphyses.²⁰ Surgery may be required later to treat such disturbances.

This study demonstrates the feasibility of closure of the PDA using venous access alone. In a consecutive series of 179 patients evaluated at our institute, we were able to close PDAs in 151 patients with venous access alone. The key steps in our opinion, included, careful patient selection by thorough 2D and color flow imaging, angiographic definition of the duct and ampulla by hand injection within the duct ampulla or proximal descending thoracic aorta and, the use of color flow imaging to assess residual flows.

The advantages of avoiding arterial access include, avoiding heparinization thereby potentially accelerating flow elimination across the duct and elimination of inherent risks of arterial puncture (bleeding, femoral artery thrombosis). The procedure is considerably short as indicated by the short fluoroscopy time²¹ and many patients can potentially be discharged early.

The potential limitations of this approach are as follows:

• 1.A good echocardiographic window is a requirement. A poor echocardiographic window was relatively more common in older children, where accurate assessment of the duct size was challenging. In these patients, the pre procedural visualization of the duct and its sizing was adequately assessed by the angiographic measurement from the venous side. Most of these patients were older children and device position and residual flow were reasonably assessed even with echocardiography.
• 2.The need for arterial access for assessment of aortic flow in a small child with large device was needed in three patients.
• 3.An echocardiography machine needs to be available in the cardiac catheterization laboratory. This may not always be possible in all institutions.
• 4.Device embolization: In the event of device embolization, arterial access is necessary to monitor hemodynamics and may be required when attempting to retrieve the device. In case the device embolizes into the descending aorta, the pulses may be feeble and obtaining arterial access may be an issue.

5. Conclusions

PDA device closure without the use of arterial access can be performed safely and effectively. Patient selection and pre procedural planning by detailed echocardiography are essential for accomplishing device closure of PDA without arterial access. The procedure is simplified considerably and many patients can be discharged on the day of the procedure. The procedure is relatively straight forward in older children and small to moderate sized ducts in smaller children. This technique is also feasible in carefully selected infants with large ducts. With increasing experience, it may be possible to occlude the PDA of small children with large ducts without arterial access.

Conflicts of interest

All authors have none to declare.

Appendix A. Supplementary data

The following are the supplementary data related to this article:

Video 1

Hand injection through long sheath in lateral projection delineating PDA via retrograde filling; a conical shaped structure with wide end towards aorta and tapering end (white arrow) towards pulmonary artery.

Video 2

Hand injection through long sheath in right anterior oblique view showing anatomy of PDA with narrow end (white arrow) towards pulmonary artery.

Video 3

Hand injection through long sheath in left anterior oblique view with cranial angulation after placement of device (delivery cable still attached) showing normal filling of branch pulmonary arteries.