Transcatheter Coil Occlusion of Surgical Fenestration After Fontan Operation
Background Fenestration of the Fontan circulation that results in a residual right-to-left shunt has improved operative survival rates among high-risk patients. Late closure of the fenestration by use of a transcatheter umbrella device has achieved separation of the systemic and pulmonary venous circulations, “completing” the Fontan pathway. Because use of umbrella devices is restricted, many institutions continue to perform only nonfenestrated Fontan procedures.
Methods and Results Five children 3.5 to 8.3 years old (mean, 5.1 years) underwent cardiac catheterization 0.5 to 24 months (mean, 10 months) after operation for the purpose of occluding a persistently patent Fontan fenestration. Once candidacy was determined, an 8-mm×10-cm Gianturco coil was delivered to straddle the fenestration with established techniques for coil occlusion of patent ductus arteriosus. Complete occlusion occurred in 4 of 5 patients, in 2 of the 4 before they left the catheterization laboratory. One patient had a residual angiographic shunt but had complete closure within 24 hours by echocardiography. In 1 patient who had a residual shunt at 24 hours, the fenestration was completely closed at 1 month after coil placement. One patient had residual shunting at 2 months but saturations have increased 15% to 17% since coil placement. No embolizations (early or late), clinical hemolysis, thromboembolic events, or hemodynamic deterioration occurred among patients during 1- to 14-month follow-up periods.
Conclusions A persistently patent fenestration after Fontan operation may be closed with a Gianturco coil. This universally available alternative to umbrella devices may make the fenestrated Fontan a more appealing option to centers that had not previously considered its use.
Fenestration of the Fontan circulation, leaving a controlled right-to-left shunt at surgery, has been shown to improve operative survival rates among high-risk patients and to shorten duration of hospital stay and pleural effusions.1 2 3 Late elimination of the residual shunt has achieved separation of the systemic and pulmonary venous circulations, “completing” the Fontan pathway, at the expense of a reduction in cardiac index and tissue oxygen delivery.3 4 5 6 7 8 9 Closure has been accomplished either surgically, with a snare-controlled, adjustable atrial septal defect as described by Laks et al,2 3 4 or with a transcatheter umbrella device, as first described by Bridges et al.5 6 7 8 9
Because transcatheter umbrella devices remain unavailable to all but a few pediatric heart centers, many institutions continue to perform only nonfenestrated Fontan procedures. For those centers that perform a fenestrated Fontan operation, residual shunts have been left open pending device availability. For residually patent defects that require closure, we report a new approach: transcatheter occlusion of the patent fenestration with a Gianturco coil (Cook Inc).
Five patients who had been palliated previously by fenestrated Fontan operations (between April 1993 and October 1995) underwent cardiac catheterization 10 days to 24 months (mean, 10 months) after operation to evaluate candidacy for coil occlusion of a patent fenestration. Informed consent was obtained in each case for test occlusion and for coil occlusion of the fenestration.
All fenestrated Fontan operations involved a total cavopulmonary anastomosis with direct superior vena cava–to–pulmonary artery connection and extracardiac baffling of the inferior vena cava (or hepatic veins) to the pulmonary artery. Pathway fenestration was accomplished by creation of a 4.0-mm defect in the lateral atrial wall to allow direct shunting from the external systemic venous chamber to the pulmonary venous atrium. All patients were treated with aspirin from the second postoperative day.
At catheterization while patients were under intravenous sedation, femoral venous and arterial access was obtained in 4 of 5 patients; in the fifth patient, who had an interrupted inferior vena cava, the right internal jugular vein was used for access. A complete hemodynamic assessment of the Fontan pathway was made in each patient. Systemic venous saturations were checked throughout the pathway to rule out residual left-to-right shunts. Cardiac index was estimated by use of measured oxygen consumption and the measured oxygen-content difference between arterial and mixed systemic venous samples (270 CO-Oximeter, Ciba-Corning). Tissue oxygen delivery (the product of arterial oxygen content and cardiac index) was calculated.8 Systemic venous pressure was measured in all limbs of the Fontan pathway to identify pathway gradients. A balloon-tipped catheter was passed across the fenestration, and the fenestration was test occluded as described previously.6 7 8 9 Systemic venous pressure, cardiac index, and tissue oxygen delivery were measured again after 10 minutes of test occlusion. Fenestration occlusion was undertaken4 7 8 if there was no significant increase in systemic venous pressure or significant decrease in cardiac index or tissue oxygen delivery.
Technique of Coil Occlusion
The approach to coil delivery was based on current techniques of transcatheter coil occlusion of patent ductus arteriosus.10 11 Fenestration diameter was estimated by use of a systemic venous angiogram (Fig 1).⇓ Coil-loop diameter was selected to be at least twice fenestration diameter, and coil length was chosen to allow for four full coil loops. In each case, a 0.038-in coil 8 mm in diameter by 10 cm long was used for occlusion. A coil-delivery catheter was passed through the fenestration to the pulmonary venous side of the baffle. Two full loops of coil were delivered to the pulmonary venous atrium and two to the systemic venous side of the fenestration. A repeat angiogram was performed 10 minutes after coil implantation. All patients received a continuous heparin infusion (25 U·kg−1·h−1) overnight and were discharged on aspirin therapy. All patients underwent Doppler evaluation of fenestration within 24 hours after catheterization.
Table 1⇓ lists underlying diagnoses, previous operations, and ages at Fontan procedure and catheterization for each patient. During test occlusion of the fenestration, the increase in systemic venous pressure (range, 0 to 2 mm Hg) and decreases in cardiac output (range, −13% to −34%) and tissue oxygen delivery (range, 1% to −17%) were judged acceptable to proceed with fenestration closure in all 5 patients (Table 2⇓). All patients experienced a significant increase in measured systemic arterial saturation (range, 7% to 20%; P<.01) during test occlusion. Fenestration diameter (estimated angiographically) ranged from 2.6 to 3.2 mm.
After coil implantation, there was an immediate increase in systemic saturation, similar to that seen with temporary balloon occlusion. Arterial saturation after coil placement ranged from 91% to 96% (room air). With administration of oxygen by nasal cannula, PO2 ranged from 71 to 295 mm Hg. Angiography was performed 10 minutes after coil placement (Fig 2⇓) and demonstrated complete fenestration closure in 2 patients (patients 1 and 3) and small residual leaks in 2. One of the latter patients, patient 4, had complete closure as assessed by Doppler within 24 hours of the procedure. The other, patient 5, had a trivial residual shunt at 24 hours but had complete fenestration closure as assessed by Doppler at 4 weeks. Patient 2 developed upper airway obstruction after coil implantation, and the procedure was terminated prematurely (before angiography). Follow-up echocardiography the next day showed a residual but markedly reduced right-to-left shunt. This patient had systemic saturations of 74% before coil placement and 91% (PO2=71 mm Hg) on room air after coil implantation. Saturation ranged from 89% to 92% at late follow-up.
All patients remain free of late coil migration (as assessed by echocardiography), clinically apparent thromboembolic events, hemolysis, and hemodynamic deterioration at 1 to 14 months of follow-up.
Since fenestration of the Fontan pathway is effective for reducing perioperative morbidity and mortality rates in high-risk patients,1 2 3 our center and others have elected to leave a surgical fenestration in all Fontan operations.
After operation, patients with a fenestration can be divided clinically into three groups. The first subset of patients has excessive right-to-left shunting in the perioperative period secondary to either high resistance through the Fontan circuit or insufficient resistance at the fenestration. These patients may require early intervention to correct residual pathway obstruction, to eliminate aortopulmonary collaterals, or to close or reduce the size of the fenestration.
A second subset has spontaneous closure of the fenestration. At our institution, 4 of 14 patients who underwent a fenestrated Fontan procedure during the same time period had spontaneous closure 1 day to 3 months after surgery.
The remaining patients who have a persistently patent fenestration are mildly cyanotic but clinically well. The “correct” management of this group is controversial. The risks of maintaining fenestration patency include the following.
1. Several centers have demonstrated a significant incidence of thrombus formation within the Fontan pathway,12 13 which places involved patients at an increased but undefined risk of systemic thromboembolic events as long as the residual right-to-left shunt is present. These children may be treated prophylactically with warfarin, but the benefit of such an approach is not known. Achieving a therapeutic international normalized ratio is a difficult task in young patients. The long-term bleeding risks of warfarin in young, active children likewise is unknown.
2. These children can increase the right-to-left shunt with exertion, which may cause them to become more cyanotic and thus exercise intolerant.
3. The chronic effects of persistent mild cyanosis on the developing child are poorly defined.
Closure of the defect in the catheterization laboratory, however, consistently has been shown to cause an acute reduction in cardiac index and tissue oxygen delivery, which has led some authors to speculate that there may be hemodynamic benefit to maintaining fenestration patency.8 This decrease in cardiac index is due in part to the acute reduction in ventricular preload that occurs with fenestration occlusion. No data are available to determine whether this acute effect and the chronic, long-term hemodynamic effects of fenestration closure are equivalent. In our small group, all patients remain asymptomatic after fenestration closure and have increased activity levels according to parental reports.
Although the advantages and disadvantages of fenestration closure can be discussed only in the abstract, in practice, centers with access to the umbrella device continue to close these defects.
This small series demonstrates the feasibility of use of a Gianturco coil to occlude the persistently patent fenestration. In addition to its availability, the coil has several theoretical advantages over the umbrella device for fenestration closure: (1) The coil is smaller than umbrella devices. This may lessen the potential for systemic and pulmonary venous pathway obstruction, particularly in small patients. (2) The thrombogenic surface of the coil or umbrella on the pulmonary venous side of the circulation poses an ongoing risk of systemic thromboembolization unrelated to right-to-left shunting. The smaller coil may reduce that risk. (3) The coil is inexpensive compared with the umbrella device.
We used a snare (Microvena Corp) in our first 2 patients to “grab” the systemic venous end of the coil as it was released to try to minimize the risk of systemic coil embolization. Conceptually, this was a useful approach; however, it was time-consuming, required additional venous access, and was ultimately unnecessary in the last 3 patients. As it is further developed, the new detachable coil system (Cook Inc) may be useful.
With complete fenestration closure as the goal of this intervention, our coil closure results are comparable to the only series in which residual shunts were reported after umbrella placement.9 In that group, 10 of 12 fenestrations were completely occluded. With the coil, there has been complete closure and normalization of systemic arterial saturation in 4 of 5 patients and a marked increase in systemic arterial saturation in the fifth. Had the procedure on this fifth patient not been interrupted, placement of a second coil might have resulted in complete fenestration closure. If complete closure does not occur, recatheterization may be possible at a later date.
The collaboration of cardiothoracic surgeons and interventional cardiologists to improve long-term outcomes has advanced the science and management of congenital heart disease. The fenestrated Fontan procedure is one such example. Disseminating these approaches is difficult without widespread access to the necessary technology. In contrast to the umbrella device, the equipment required to perform coil occlusion of a persistently patent Fontan fenestration is universally available. Most pediatric interventionalists already use a similar technique for occlusion of patent ductus arteriosus. The ability to close a residual right-to-left shunt after surgery when indicated may make the fenestrated Fontan operation a more appealing option to institutions that previously had not considered its use. This study is too small and follow-up is insufficient to assess long-term safety and late closure rates. However, our initial positive experience with coil occlusion of persistently patent Fontan fenestrations warrants further investigation in a larger cohort of patients.
- Received April 8, 1996.
- Revision received May 31, 1996.
- Accepted June 6, 1996.
- Copyright © 1996 by American Heart Association
Bridges ND, Mayer JE Jr, Lock JE, Jonas RA, Hanley FL, Keane JF, Perry SB, Castaneda AR. Effect of baffle fenestration on outcome of the modified Fontan operation. Circulation.. 1992;86:1762-1769.
Mavroudis C, Zales VR, Backer CL, Muster AJ, Latson LA. Fenestrated Fontan with delayed catheter closure: effects of volume loading and baffle fenestration on cardiac index and oxygen delivery. Circulation. 1992;86(suppl II):II-85-II-92.
Hijazi ZM, Fahey JT, Kleinman CS, Kopf GS, Hellenbrand WE. Hemodynamic evaluation before and after closure of fenestrated Fontan: an acute study of changes in oxygen delivery. Circulation.. 1992;86:196-202.
Bridges ND, Lock JE, Castaneda AR. Baffle fenestration with subsequent transcatheter closure: modification of the Fontan operation for patients at increased risk. Circulation.. 1990;82:1681-1689.
Lloyd TR, Fedderly R, Mendelsohn AM, Snadhu SK, Beekman RH. Transcatheter occlusion of patent ductus arteriosus with Gianturco coils. Circulation.. 1993;88:1412-1420.
Rosenthal DN, Friedman AH, Kleinman CS, Kopf GS, Rosenfeld LE, Hellenbrand WE. Thromboembolic complications after Fontan operations. Circulation. 1995;92(suppl II):II-287-II-293.