Percutaneous Transcatheter Closure of Congenital Coronary Artery Fistulae With Patent Ductus Arteriosus Occluder in Children: Focus on Patient Selection and Intermediate-Term Follow-Up Results

Abstract: Background. The prognostic implications of clinically silent coronary artery fistula (CAF) and its intermediate/long-term outcomes after transcatheter closure have not been well studied, especially in children. Aims. This study intended to determine the prognostic implications of asymptomatic CAF and to evaluate the intermediate follow-up outcomes following transcatheter occlusion with patent ductus arteriosus (PDA) in children. Methods. Eighteen children with congenital CAF were divided into two groups: the intervention group (n = 14; maximal coronary artery diameter [MCD] ≥5 mm and/or fistulous orifice diameter [FOD] >2 mm), and the non-intervention group (n = 4; MCD <5 mm and FOD ≤2 mm). Patients in the intervention group received percutaneous occlusion with PDA occluder. Clinical outcomes and follow-up data were analyzed and compared between different groups. Results. Patients with clinically silent CAF were followed for 8-130 months. At the mean follow-up of 36 months, patients in the non-intervention group did not show any changes in the measured parameters over time. In contrast, patients in the intervention group showed significant increase of MCD (8.31 ± 2.16 mm to 12.75 ± 3.01 mm; P=.001) and FOD (3.75 ± 3.42 mm to 4.82 ± 1.81 mm; P=.03). In addition, 3 cases of aneurysm formation and 2 cases of mild heart failure were detected before the patients received the attempted transcatheter closure. A total of 14 patients underwent cardiac catheterization with an attempt to close the CAF. Placement of occlusion devices succeeded in 10 patients (71.5%) and failed in 4 patients (28.5%). Ten children with successful transcatheter closure were followed 3-62 months (median, 36 months). At the medial time of 36 months, all patients with closure were in New York Heart Association functional class I and asymptomatic. The MCD decreased from 9.66 ± 3.86 mm to 7.82 ± 3.83 mm (P=.36). Conclusion. All asymptomatic CAFs in children with MCD ≥5 mm and/or FOD >2 mm should be closed as early as possible to prevent later complications. Transcatheter closure of CAF using the PDA occluder is an effective and safe approach in appropriately selected children and showed favorable intermediate-term follow-up outcomes.

J INVASIVE CARDIOL 2014;26(7):339-346

Key words: congenital coronary artery fistula, transcatheter closure, patent ductus arteriosus occluder, children, follow-up

__________________________________________

Coronary artery fistula (CAF) is an abnormal connection/passageway between a coronary artery and a cardiac chamber or a major intrathoracic vessel. Most CAF patients are asymptomatic in childhood, because fistulae are small and do not compromise myocardial blood flow. While spontaneous closure occurs in very small fistulae, most fistulae enlarge progressively and need operative repair, either by transcatheter or surgical operation. Also, because of natural progression of size over the time and increasing risk of thrombosis, endocarditis, or rupture, monitoring and early closure are recommended to prevent complications such as heart failure, infective endocarditis, myocardial ischemia, and dysrhythmias.¹ The 2011 American Heart Association (AHA) indications for cardiac catheterization and intervention in pediatric cardiac disease recommend closure of all symptomatic CAFs (class I, evidence B). Transcatheter occlusion is reasonable for the management of patients with moderate or large CAFs without clinical symptoms (class IIa, evidence C), but it is not indicated for patients with clinically insignificant CAFs, eg, normal-sized cardiac chambers (class III, evidence C).² Traditionally, surgical closure was the main procedure for CAF repair. However, a number of transcatheter studies using various devices have been reported since the first report for successful transcatheter closure by Reidy in 1983.^3-9 Furthermore, the recently developed percutaneous closure technique makes this approach an effective and safer alternative to surgery, with the advantages of a less invasive procedure, ie, shorter hospital stay and quicker recovery.¹⁰ However, the prognostic implications of clinically silent CAF and the intermediate/long-term outcomes following transcatheter closure have not been well studied, especially in children,^11-16 which makes it difficult to recommend management points and to predict the prognosis of the procedure. In the current study, we intended to determine the prognostic significance of asymptomatic CAF and to evaluate mid-term follow-up outcomes of CAF closure following transcatheter patent ductus arteriosus (PDA) occluder occlusion in children.

Methods

Patient selection. Between July 2007 and September 2012, a total of 18 children with congenital CAF were confirmed by coronary angiography in the department of pediatric cardiology of West China Second University Hospital. All of these patients were asymptomatic when they were first diagnosed. According to the final treatment, they were divided into two groups: the intervention group (n = 14; maximal coronary artery diameter [MCD] ≥5 mm and/or fistulous orifice diameter [FOD] >2 mm) and the non-intervention group (n = 4; MCD <5 mm and FOD ≤2 mm). Patients in the intervention group met the criteria for percutaneous occlusion, including the presence of symptoms related to CAF, coronary dilation (maximal coronary artery diameter ≥5 mm), and/or aneurysmal formation. Occlusion devices were successfully placed in 10 of 14 patients. In addition, according to the location, CAFs were classified as proximal or distal type in this study. The proximal type of fistula refers to those that arise near the origin of the coronary artery, and the distal type has its origin near the distal end of a coronary artery branch.^1,13 All available clinical details, cardiac catheterization data, echocardiography, and multidetector computed tomography (MDCT) parameters before closures as well as during follow-up were reviewed. Before transcatheter closure, 12 of 18 patients were followed clinically by electrocardiogram (ECG) and echocardiography.

Ethics statement. This study was approved by the University Committee on Human Subjects at Sichuan University and informed written consent was obtained from all parents/guardians of the patients prior to cardiac catheterization and angiography.

The device. The PDA occluder used in the current study is a self-expandable cone-shaped device made from nitinol wire mesh. The device has been widely demonstrated to be successful in PDA occlusion in China. The PDA occluder requires a small delivery sheath and has the ability to reposition a device until release and to recapture a misplaced device (Lifetech Scientific ShenZhen Company, Ltd). The device has a thin retention disk, which is 4-6 mm larger than the waist diameter and allows the skirt of the device to be anchored in an “elbow” or narrow point of a fistulous vessel. It is manufactured in diameters from 4-22 mm (in 2 mm increments), with a length of 7-10 mm. It can be delivered through standard 5-12 Fr delivery sheaths. The shape and sizes of the PDA occluder and the French sizes of corresponding delivery sheaths are shown in Figure 1.

Transcatheter closure procedure. The procedure was performed under general anesthesia. Heparin was administered (100 U/kg) after the femoral venous and arterial accesses were established. After analysis of hemodynamic data, aortic root angiography and selective coronary angiography were performed to assess the anatomy and dimensions of the coronary arteries and fistulae. The PDA occluder was chosen for transcatheter closure on the basis of the anatomy and location of the CAF. The general guidance for PDA occluder selection is that the occluder waist diameter is approximately two times the diameter of the narrowest segment of the fistula that is most proximal to the fistulous orifice or the fistulous orifice (the communication between the fistula and the cardiac cavities). After a Judkins catheter was intubated into the affected coronary artery, a coronary super-smooth guidewire (260 mm, 0.035˝; Terumo Corporation) was advanced into the affected coronary artery and through the fistula. The guidewire tip was then snared using a microvena snare catheter advanced antegradely into the superior vena cava (SVC) or the pulmonary artery (PA). Thereafter, the wire was externalized via femoral venous access. After the arteriovenous wire loop was established, a long sheath (Lifetech Scientific ShenZhen Company, Ltd) was advanced antegradely into the fistula (Figures 1 and 2). To avoid occlusion of proximal coronary branches, the device was placed at the narrowest segment, which is the fistulous orifice or mostly proximal to fistulous orifice. This is particularly important for the distal type of CAF. Before deploying the device, ECG was monitored for up to 10 minutes to ensure no myocardial ischemia. Repeat angiography was performed in the affected coronary artery and ascending aorta to confirm that the device was in an appropriate position and complete occlusion was achieved without occlusion of an adjacent coronary branch. Thereafter, the device was released. Following the procedure, all patients were put on oral aspirin 5 mg/kg/day for 6 months.

Follow-up protocol. Children undergoing successful CAF closure (n = 10) were subjected to 48 hours of dynamic ECG monitoring, as well as a 12-lead ECG and echocardiography at 24 hours post procedure. In addition, serum troponin levels after CAF closure were also measured as another reference for cardiac ischemia. After hospital discharge, patients were followed clinically as well as with chest radiography, 24-hour dynamic ECG, and echocardiography at 1, 3, 6, and 12 months during the first year and annually thereafter. The MDCT coronary angiography after CAF closure was performed if available. The MCD was measured in the parasternal short-axis view. The fistulous orifice was detected using color Doppler imaging. All measurements were obtained in diastolic frame. Each value was taken from the average of three measurements. Data with respect to MCD and FOD of all children before closures as well as during follow-up were collected by the same ultrasound doctor. Children with clinically silent CAFs (n = 12) were followed clinically with ECG and echocardiography since first diagnosed.

Data analysis. Quantitative data were expressed as mean ± standard deviation. Paired t-test was used to compare the differences between two groups. It was considered statistically significant when P<.05.

Results

Clinical and echocardiographic follow-up outcomes of 12 patients with clinically silent CAF. Patients with clinically silent CAF were followed up to 8-130 months. At the mean time of 36 months, patients in the non-intervention group (n = 4) did not show any changes in the measured parameters over the follow-up period: MCD (3.55 ± 3.42 mm to 3.39 ± 0.17; P=.17) and FOD (1.75 ± 0.50 mm to 1.50 ± 0.58 mm; P=.39). In contrast, patients in the intervention group showed significant increase of MCD (8.31 ± 2.16 mm to 12.75 ± 3.01 mm; P=.001) and FOD (3.75 ± 3.42 mm to 4.82±1.81 mm; P=.03) over the 36 months. In addition, aneurysm formation and mild heart failure were detected in 3 and 2 patients, respectively, before they received the attempted transcatheter closure. The detailed follow-up data are summarized in Table 1.

Clinical characteristics of the patients undergoing an attempted CAF closure. As shown in Table 2, a total of 14 patients (9 males; age, 19 to 135 months; median age, 62.5 months; weight, 17.89 ± 7.08 kg) underwent cardiac catheterization with an attempt to close the CAF. Occlusion devices were successfully placed in 10 patients (71.5%). The attempted closure failed in 3 patients (Table 2; cases #11, #12, and #14) because of high degree of vessel tortuosity and/or presence of large aneurysm that prevented a successful establishment of an arteriovenous loop. The attempted CAF closure failed in another case (#13), because a coronary branch was so close to the fistulous orifice (Figure 3) that the device placement might have led to myocardial ischemia. Eventually, cases #11, #12, and #14 underwent surgical repair and case #13 was only followed regularly by ECG and echocardiography without further operation, at his parent’s insistence. 

The most common origin of CAF is from the right coronary artery (RCA; 10/14 [71.4%]), followed by the left anterior descending coronary artery (LAD; 2/14 [14.3%]), left main coronary artery (LM; 1/14 [7.1%]), and left circumflex coronary artery (LCX; 1/14 [7.1%]). All of them drained into the right heart (8 into the ventricle and 6 into the atrium). The mean cardiothoracic ratio (C/T) of the patients was 0.59 ± 0.02 mm. The mean MCD of all patients was ≥5 mm (9.66 ± 3.86 mm) and the mean FOD was >2 mm (3.75 ± 1.37 mm) before the procedure. Clinical characteristics of the patients undergoing attempted CAF closure are summarized Table 2. 

Immediate outcome following device closure. The hemodynamic and angiographic characteristics of patients undergoing successful transcatheter closure are summarized in Table 3. All of the CAFs were successfully occluded antegradely with the PDA occluder using arteriovenous wire loop approach. The diameter of the narrowest segment of fistulae (NSF) was 3.53 ± 1.09 mm. The size of deployed devices was 4/6 (for 3 cases), 6/8 (for 2 cases), 8/10 (for 2 cases) and 10/12 (for 2 cases). The procedure duration was 101.5 ± 38.66 minutes. 

Repeated angiograms following device deployment indicated that complete CAF closure was achieved in 9 patients, except that case #5 had minimal residual shunt, which has a secondary drainage site from LM to right atrium. After a rigorous evaluation, a conservative treatment was adopted. Case #8 had transient ST-T wave changes during the procedure. Case #3 had a persistent incomplete right bundle branch block (IRBBB) and sinus bradycardia for 5 days, which was corrected with dexamethasone (1 mg/kg) and atropine (0.01-0.03 mg/kg) prior to discharge. Dorsal arterial pulse disappeared for 1 hour in case #7 after the procedure, which might be related to the long procedure duration. Urokinase (300,000 U) was given to this patient and no symptom of thrombus formation was found. There were no immediate deaths or cases of myocardial ischemia. 

Intermediate-term follow-up results after CAF closure. At the median time of follow-up (36 months; range, 3-62 months), all patients undergoing successful transcatheter closure were clinically asymptomatic and free from murmurs, and were in New York Heart Association (NYHA) functional class I. The latest chest radiography and echocardiography, available in all patients, indicated complete occlusion in all patients except for case #5, whose secondary drainage was still present. The cardiothoracic ratio decreased significantly from 0.59 ± 0.02 to 0.55 ± 0.02 (P=.001). In addition, the MCD decreased from 9.66 ± 3.86 mm to 7.82 ± 3.83 mm, although failing to reach statistically significant level (P=.36). During the follow-up, 2 patients received MDCT coronary angiography and neither recanalization nor thrombus formation were found. There were no deaths or other significant complications, such as chest pain or fistula dissection. Clinical and echocardiographic profiles for intermediate-term follow-up after transcatheter closure of CAF are summarized in Table 4.

Discussion

Coronary artery fistula is considered a major coronary anomaly. It is present in 0.002% of the general population and accounts for 0.4% of all cardiac malformations.¹⁷ Most patients with CAF are asymptomatic and most CAFs are encountered as incidental findings during coronary angiography. All pediatric patients in the current study were first referred to our hospital because of heart murmurs incidentally heard during routine physical examinations. Selective closure of symptomatic CAF in childhood has been advocated to prevent later complications. However, the best management strategy for asymptomatic CAF is still controversial, since both spontaneous closure and life-threatening complications have been reported.¹ The guidelines recommend closure of all moderate and large fistulae without clinical symptoms to prevent future complications, whereas closure is not indicated for patients with small and insignificant CAF. However, there is no clear definition of “large,” “moderate,” and “small” in terms of CAF size.² In addition, there is a paucity of information regarding the prognostic implications of incidentally identified, clinically silent CAF.^15,16 With the absence of data regarding follow-up outcomes, it is difficult to make recommendations for management strategies and optimal intervention timing. In the present study, all patients were asymptomatic when they were first diagnosed with CAF. Our follow-up data indicated that while patients in the non-intervention group remained asymptomatic and their MCD and FOD had no changes, patients in the intervention group showed gradually increased MCD and FOD, aneurysm formation, and symptoms of mild heart failure. It was suggested that if pediatric patients are categorized into intervention group, their CAFs have a tendency to enlarge and have complications later. Valente et al concluded that greater fistula length, diameter, and degree of tortuosity result in a higher incidence of thrombus formation after CAF closure.¹⁸ In addition, older age at intervention was another risk factor for thrombus formation after CAF closure.^1,13 In light of the findings in the present study and previous studies, we advocate the closure of all CAFs with MCD ≥5 mm and/or FOD >2 mm as early as possible for asymptomatic CAF pediatric patients. For children with MCD <5 mm and FOD ≤2 mm, a conservative treatment may be a relatively better choice because those CAFs are hemodynamically insignificant and even tend toward spontaneous closure.^15,16

Percutaneous transcatheter closure is an emerging alternative for patients with suitable anatomy and absence of other complex cardiac malformations. Previous studies have reported successful cases of acute procedure using a variety of devices, including coils, detachable balloons, covered stents, the Amplatzer ductal occluder (ADO), and the Amplatzer vascular plug (AVP).^3-7,11,19-28 Coils were used primarily in smaller CAFs, offering the advantages of smaller sheath and delivery catheter sizes, as well as lower cost. However, the coil option is technically demanding and time consuming for patients with moderate or large fistulas and many have a residual shunt.^4,11 All patients in our study had a moderate or large CAF, and thus coils were not selected. Successful occlusion of large fistulas with ADO and AVP has been reported. These two detachable devices have several advantages, including a high success rate of complete occlusion and relatively easy implantation procedure.^{6,7,20,24-26,28} The AVP is occasionally used for CAF closure, but it lacks retention discs and polyester fabric, has poor stability, and frequently leaves a significant residual shunt as well as the risk of embolization if released.^7,28 Therefore, the ADO seems to be a better choice for moderate and large CAFs. The PDA occluder used in the present study is very similar to the ADO, and thus has all the advantages of the ADO. Our device choice was also more economically feasible, because it is manufactured in our home country of China, and its cost is more affordable for most of our patients.

Our data indicated that most CAFs with drainage to a right-sided heart chamber could be successfully occluded with a PDA device using an arteriovenous wire loop approach. However, in our study, the attempted closure failed in 3 patients due to high-degree tortuosity and large aneurysm and failed in 1 patient due to the presence of a coronary branch next to a fistulous orifice. It was suggested that extra caution is needed for catheter techniques when a patient has the above-mentioned conditions. In addition, it is definitely vital to determine whether acute myocardial infarction occurs following device placement. Based on our successful experience, we recommend the following considerations to avoid postprocedure ischemia. First, it is important to identify the type of CAF (proximal or distal) using aortic root angiography and selective coronary angiography. Our data imply that patients with proximal CAF are at relatively lower risk for adverse events from myocardial ischemia or infarction after closure, since there are no normal coronary branches arising from the fistula. For patients with distal CAF, extreme care should be taken to assess whether there are any visible coronary arteries adjacent to the fistulous orifice. To avoid occlusion of unidentified distal coronary branches, a 6-limb ECG was monitored for up to 10 minutes to assess for myocardial ischemia before deploying the device. Second, following device deployment, repeat angiography for the affected coronary artery and ascending aorta is also necessary to ensure there is no occlusion in adjacent coronary branches. Last, even after the device is placed at the fistula drainage site, it may still lead to myocardial infarction in a small area. Therefore, it is necessary to monitor patient for ST-T changes as well as the serum troponin levels after CAF closure for 48 hours.

Regarding the efficacy and safety of transcatheter closure of CAFs, many studies conducted in adult patients have been published in the literature.^{4,12-14,18,29-32} The reported complications include device migration, recanalization of the fistula, coronary thrombus formation, and myocardial ischemia or infarction. While the data from adult patients may serve as a good reference for pediatric patients, the outcome of CAF closure in children should be carefully evaluated as a separate group, because of their physical growth and resulting anatomic changes. However, very limited information is available for intermediate- and long-term outcomes after transcatheter closure in pediatric patients. The current study evaluated the intermediate follow-up outcomes of one of the largest transcatheter closure series on the PDA occluder in children. No device migration or recanalization were observed during our follow-up period. We attribute this success to the correct choice of device size. Because vascular elasticity is relatively better in children, the size of PDA occluder we chose (the small waist diameter) was roughly twice the diameter of the narrowest segment of the fistula or the fistulous orifice. 

One concern following CAF closure was thrombus formation in the stump, which may result in occlusion of distal coronary branches and myocardial ischemia or infarction. Previous studies have implied that large distal CAFs may possess higher risk for coronary thrombosis than the proximal type of CAF.¹³ To date, there is no generally accepted therapeutic approach for large distal-type CAFs due to the lack of data for long-term follow-up outcomes. In the present study, half of our patients with CAF closure had large distal-type CAFs and all of them were asymptomatic during the follow-up period. In addition, the MDCT available in 2 patients did not show signs of thrombosis formation. Valente et al proposed that older age at the time of intervention (>20 years old) was associated with increasing chance of major adverse outcome post occlusion, including thrombosis.¹⁸ It was partially attributed to the increase of comorbid factors in adult patients, such as smoking, hypertension, hyperlipidemia, obesity, diabetes, and hypercoagulable disorders, which may hasten the process of procoagulation and thrombosis.^12-14,18,33 Unlike the previous studies, patients in the current study were much younger and are unlikely to have adult risk factors. Although coronary artery dilatation still persisted after a mean of 36 months, decrease of the MCD was achieved in all patients. We speculated that closure of CAFs at a younger age may provide a more favorable condition to allow the dilated coronary artery to restore the normal size, and thereby the risk for thrombosis may be lower than in adult patients. Several earlier studies have also concluded that the absence of antiplatelet/anticoagulation after closure may also form the basis for thrombosis formation.^13,29 All of the patients in our study were prescribed oral aspirin 5 mg/kg/day for 6 months after CAF closure, which may partly contribute to the favorable intermediate-term outcome in our study. Taken together, we would recommend that children with large distal CAFs should undergo early closure to avoid progressive enlargement of the proximal coronary artery and thereby decrease the risk for later complications. 

Encouragingly, we confirmed that greater initial size of MCD and/or FOD in asymptomatic pediatric patients with CAF is possibly related to the late occurrence of complications. Based on our experience, we recommend the closure of asymptomatic CAF in children as early as possible if MCD is ≥5 mm and/or FOD is >2 mm. In addition, based on the favorable intermediate-term follow-up outcomes, we think that transcatheter closure of CAF with a PDA occluder is an effective and safe approach in appropriately selected children with moderate and large CAFs. Furthermore, as discussed above, we suggest that closure of moderate and large distal CAFs at a younger age may produce favorable conditions for remodeling conduit coronary artery and reducing risk for thrombosis. 

Study limitations. Our study has some limitations, including the small sample size and unavailability of follow-up angiogram or MDCT in all patients. In addition, it is necessary to optimize the duration for antiplatelet/anticoagulation following CAF closure. Furthermore, management of CAF in the current study was primarily based on expert consensus as opposed to a large randomized controlled trial. Further investigation is still necessary to make definitive recommendations. Last, some severe complications, such as myocardial infarction, may occur much later after CAF closure.¹³ Therefore, multicenter longitudinal studies with more patients and long-term follow-up are still needed to better understand the best management approach for this rare but important disorder.

References

1. Mangukia CV. Coronary artery fistula. Ann Thorac Surg. 2012;93(6):2084-2092. Epub 2012 May 5.
2. Feltes TF, Bacha E, Beekman RH 3rd, et al. Indications for cardiac catheterization and intervention in pediatric cardiac disease: a scientific statement from the American Heart Association. Circulation. 2011;123(22):2607-2652.
3. Perry SB, Rome J, Keane JF, Baim DS, Lock JE. Transcatheter closure of coronary artery fistulas. J Am Coll Cardiol. 1992;20(1):205-209.
4. Armsby LR, Keane JF, Sherwood MC, Forbess JM, Perry SB, Lock JE. Management of coronary artery fistulae: patient selection and results of transcatheter closure. J Am Coll Cardiol. 2002;39(6):1026-1032.
5. Auf Der Maur C, Chatterjee T, Erne P. Percutaneous transcatheter closure of coronary-pulmonary artery fistula using polytetrafluoroethylene-covered graft stents. J Invasive Cardiol. 2004;16(7):386-388.
6. Balaguru D, Joseph A, Kimmelstiel C. Occlusion of a large coronary-cameral fistula using the amplatzar vascular plug in a 2-year old. Catheter Cardiovasc Interv. 2006;67(6):942-946.
7. Fischer G, Apostolopoulou SC, Rammos S, Kiaffas M, Kramer HH. Transcatheter closure of coronary arterial fistulas using the new Amplatzer vascular plug. Cardiol Young. 2007;17(3):283-287. Epub 2007 Apr 20.
8. Cebi N, Schulze-Waltrup N, Fromke J, Scheffold T, Heuer H. Congenital coronary artery fistulas in adults: concomitant pathologies and treatment. Internat J Cardiovasc Imag. 2008;24(4):349-355.
9. Xu L, Xu ZY, Jiang SL, et al. Transcatheter closure of coronary artery fistula in children. Chin Med J (Engl). 2010;123(7):822-826.
10. Kasar P, Kotecha M, Vimala J, Pavithran S, SureshKumar R. Management of coronary artery fistulas in current era. Ann Pediatr Cardiol. 2010;3:224-225.
11. Liang CD, Ko SF. Midterm outcome of percutaneous transcatheter coil occlusion of coronary artery fistula. Pediatr Cardiol. 2006;27(5):557-563.
12. Zhu XY, Zhang DZ, Han XM, et al. Transcatheter closure of congenital coronary artery fistulae: immediate and long-term follow-up results. Clin Cardiol. 2009;32(9):506-512.
13. Gowda ST, Latson LA, Kutty S, Prieto LR. Intermediate to long-term outcome following congenital coronary artery fistulae closure with focus on thrombus formation. Am J Cardiol. 2011;107(2):302-308.
14. Jama A, Barsoum M, Bjarnason H, Holmes Jr DR, Rihal CS. Percutaneous closure of congenital coronary artery fistulae: results and angiographic follow-up. JACC Cardiovasc Interv. 2011;4(7):814-821.
15. Sherwood MC, Rockenmacher S, Colan SD, Geva T. Prognostic significance of clinically silent coronary artery fistulas. Am J Cardiol. 1999;83(3):407-411.
16. Yilmazer MM, Demir F, Yolbas I, Bilici M. Spontaneous closure of a symptomatic coronary artery fistula just within a few days of newborn period. Congenit Heart Dis. 2014;9(1):E27-E30. Epub 2013 Apr 22.
17. Reeves ER, Jahromi BS, Ling FS. Coil embolization of a coronary artery to left ventricular fistula using detachable coils. Catheter Cardiovasc Interv. 2013;82(1):155-158.
18. Valente AM, Lock JE, Gauvreau K, et al. Predictors of long-term adverse outcomes in patients with congenital coronary artery fistulae. Circ Cardiovasc Interv. 2010;3(2):134-139. Epub 2010 Mar 23.
19. Krabill KA, Hunter DW. Transcatheter closure of congenital coronary arterial fistula with a detachable balloon. Pediatr Cardiol. 1993;14(3):176-178.
20. Thomson L, Webster M, Wilson N. Transcatheter closure of a large coronary artery fistula with the Amplatzer duct occluder. Catheter Cardiovasc Interv. 1999;48(2):188-190.
21. Boccalandro F, Awadalla H, Smalling RW. Percutaneous transcatheter coil embolization of two coronary fistulas originating from the left main ostium and left anterior descending artery. Catheter Cardiovasc Interv. 2002;57(2):221-223.
22. Fletcher S, Awadallah S, Amin Z. Extension of transcatheter coil occlusion to the treatment of complex coronary fistulas. J Interv Cardiol. 2003;16(2):165-169.
23. Gao W, Zhou AQ, Yu ZQ, Wang RF, Li F, Zhong YM. Transcatheter closure of cardiac malformation using coil implantation in children. Chin J Contemporary Pediatr. 2005;7:43-46.
24. Behera SK, Danon S, Levi DS, Moore JW. Transcathether closure of coronary artery fistulate using the Amplatzer duct occluder. Catheter Cardiovasc Interv. 2006;68(2):242-248.
25. Wiegand G, Sieverding L, Kaulitz R, Hofbeck M. Transarterial and transvenous approach for transcatheter closure of a large coronary artery fistula with the Amplatzer vascular plug. Pediatric Cardiology. 2009;30(2):172-175. Epub 2008 Jul 25.
26. Bruckheimer E, Harris M, Kornowski R, Dagan T, Birk E. Transcatheter closure of large congenital coronary-cameral fistulae with Amplatzer devices. Catheter Cardiovasc Interv. 2010;75(6):850-854.
27. Wang JK, Wu MH, Ling MT, Chiu SN. Transcatheter treatment of coronary artery fistula using a variety of devices. Cardiol Young. 2010;20:400-401.
28. Baspinar O, Kervancioglu M, Aldudak B. Retrograde transcatheter closure of large coronary fistulas with the amplatzer vascular plug in children. Cardiol Young. 2011;21(1):104-106. Epub 2010 Oct 27.
29. Kharouf R, Cao QL, Hijazi ZM. Transcatheter closure of coronary artery fistula complicated by myocardial infarction. J Invasive Cardiol. 2007;19(5):E146-E149.
30. Duan ZZ, Zhu XY, Han XM, Cui CS, Sheng XT, Wang QG. Transcatheter closure of congenital coronary artery fistulae: immediate and long-term follow-up results. Cardiology. 2009;114:21.
31. Ali N, Hung LC, Samion H. Acute myocardial ischaemia following transcatheter occlusion of coronary artery fistula in an infant: a case report. Cardiol Young. 2010;20:295.
32. Koneti NR, Penumatsa RR, Kanchi V, Yendamary SV, Arramraju SK. Transcatheter closure of coronary arterial-venous fistulae (CAVF) and management of complications. Cardiol Young. 2010;20:382-383.
33. Oto A, Aytemir K, Cil B, et al. Percutaneous closure of coronary artery fistulae in adults with intermediate term follow-up results. J Interv Cardiol. 2011;24(3):216-222. Epub 2011 Jan 31.

_____________________________________________

^*Joint first authors.

From the ¹Department of Pediatric Cardiovascular Disease, West China Second University Hospital, Sichuan University, Chengdu, Sichuan, China; ²West China Medical School of Sichuan University, Chengdu, Sichuan, China; and ³The Pulmonary Vascular Remodeling Research Unit, West China Institute of Women and Children’s Health, West China Second University Hospital, Sichuan University, Chengdu, Sichuan, China.

Funding: The present study was supported by the National Science Fund of China (Grant No.81270226 and 81070136)

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.

Manuscript submitted April 25, 2013, provisional acceptance given July 3, 2013, final version accepted November 20, 2013.

Address for correspondence: Professor Yimin Hua, Department of Pediatric Cardiovascular Disease, West China Second University Hospital, Sichuan University, No. 20, Section 3, RenminNanLu Road, Chengdu, Sichuan, 610041, China. Email: nathan_hua@126.com