Introduction
Coronary artery fistula is a rare condition characterised by an abnormal connection between coronary arteries and a cardiac chamber or large vessel, bypassing the capillary network, which occurs in about 0.002% of the general population. Reference Stout, Daniels and Aboulhosn1 Among the patients undergoing coronary CT angiography, the prevalence of coronary artery fistulas has been reported as 0.9%, with coronary artery to pulmonary artery fistulas account for 76.8% of the cases. Reference Lim, Jung, Lee and Lee2 This report presents a rare case of symptomatic coronary artery to pulmonary artery fistulas in a Fontan physiology patient, successfully treated with vascular plugs placed from the pulmonary side.
Case
A 32-year-old woman with a complex cardiac history presented with dyspnoea. She was diagnosed with pulmonary atresia with an intact ventricular septum, accompanied by a bicuspid aortic valve at birth. At 3 months old, she underwent a left modified Blalock-Taussig-Thomas shunt, followed by a right modified Blalock-Taussig-Thomas shunt at 1 year. Although she had been followed up for several years with consideration for biventricular repair, right ventricle decompression was abandoned due to a very small tricuspid valve, bipartite right ventricular anatomy, sinusoidal communication, and muscular pulmonary atresia. At 11 years old, she was referred to our hospital to undergo a bidirectional Glenn procedure, aortic valve replacement with a 23 mm CarboMedics prosthesis (CarboMedics Inc., Austin, TX), and ligation of the bilateral Blalock-Taussig-Thomas shunts. A year later, she had an extracardiac Fontan operation. At 31, she was diagnosed with liver cirrhosis (Child-Pugh Class B) complicated by chronic ascites. In addition, she had experienced multiple episodes of atrial flutter and fibrillation, which were managed with amiodarone. At age 32, she was diagnosed with symptomatic aortic stenosis caused by dysfunction of her mechanical aortic valve. Consequently, she underwent re-replacement of the aortic valve with a 25 mm On-X mechanical valve X (On-X Life Technologies, Austin, TX). Although the surgery was successful, she continued to experience dyspnoea on exertion and was hospitalised for further evaluation 8 months post-operation.
Upon admission, the patient was in NYHA Class II heart failure, with complaints of abdominal and leg swelling. Her medications included furosemide, spironolactone, enalapril, warfarin, aspirin, carvedilol, tadalafil, and amiodarone. She was 144 cm tall and weighed 42 kg. Her baseline oxygen saturation was 91%. A chest X-ray showed mild cardiomegaly with a cardiothoracic ratio of 59% and chronic right pleural effusion. Transthoracic echocardiography demonstrated a well-functioning mechanical valve, mild mitral regurgitation, and a left ventricular ejection fraction of 55%. A CT angiogram revealed multiple microvessels around the coronary arteries and the pulmonary artery trunk (Figure 1) .
Right and left heart catheterisation showed elevated central venous pressure of 20 mmHg and pulmonary capillary wedge pressure of 14 mmHg. There was a significant oxygen step-up in the remnants of the main pulmonary artery, with oxygen saturation of 91% in the pulmonary trunk and 52% in the superior vena cava. Angiography identified coronary to pulmonary artery micro-fistulas, and a patent left Blalock-Taussig-Thomas shunt was confirmed (Figure 2a) . Balloon occlusion of the main pulmonary artery revealed a reduction of central venous pressure to 16 mmHg. Therefore, the main pulmonary artery measuring 12 mm diameter was closed with two 16 mm Amplatzer vascular plug II (Abbott Cardiovascular, Chicago, IL) (Figure 2d). Then, a 3 mm Blalock-Taussig-Thomas shunt was also closed with a 4 mm Amplatzer vascular plug IV (Abbott Cardiovascular, Chicago, IL) (Figure 2e). Following the procedure, she continued aspirin and warfarin and was discharged the next day without complications. At follow-up, her symptoms improved, and she was able to return to work. A year later, her pleural effusion had resolved, and follow-up catheterisation showed improved hemodynamic parameters, including central venous pressure of 13 mmHg and pulmonary capillary wedge pressure of 8 mmHg. The patient’s condition had significantly improved, demonstrating positive outcomes from the intervention.
Discussion
In postoperative Fontan patients, symptoms like fatigue can result from various conditions such as Fontan pathway obstruction, protein-losing enteropathy, heart failure, arrhythmias, or elevated central venous pressure due to residual systemic to pulmonary artery shunts. Diagnosing these shunts is challenging with echocardiography alone and requires advanced imaging like CT angiography or cardiac catheterisation. In this case, angiography revealed coronary to pulmonary artery fistulas and a residual left Blalock-Taussig-Thomas shunt, which were contributing to complications in the Fontan circulation.
Coronary to pulmonary artery fistulas are often asymptomatic in normal hearts, Reference Karazisi, Eriksson and Dellborg3 with few reports in post-Fontan patients. In Fontan physiology, systemic to pulmonary artery flow adds a burden on both the cardiac load and pulmonary circulation, although the clinical significance remains debated. Some studies suggest that these fistulas may lead to pleural effusions and other complications, Reference Grosse-Wortmann, Drolet and Dragulescu4,Reference Osawa, Schaeffer and Kristina5 while others report no association with hemodynamic changes. Reference McElhinney, Reddy and Tworetzky6,Reference Bradley, McCall, Sistino and Radtke7 In this case, the decision to close the shunts was based on the possibility that they were contributing to the patient’s symptoms, as the occlusion test showed a reduction in central venous pressure.
Most coronary to pulmonary artery fistulas drain into the main pulmonary trunk, similar to this case. Reference Kim, Jung and Chun8 Generally, coronary to pulmonary artery fistulas are closed percutaneously using detachable coils via the antegrade approach. Reference Liu, Zhang, Qi, Fan and Ge9 Although it is difficult to approach from the pulmonary artery side in the normal circulation, approaching via the coronary artery involves the risk of coronary artery injury or thrombus formation. Furthermore, for multiple micro-fistulas, achieving complete closure can be practically impossible. In our case, the fistulas were successfully closed by blocking the main pulmonary trunk, which was unnecessary for the pulmonary circulation and used as a good landing zone. Although this approach might not be suitable for all Fontan patients, it is an effective option for closing complex coronary to pulmonary artery fistulas with a drainage site in a blind-end space.
Conclusion
The case highlights the importance of detecting residual systemic to pulmonary artery shunts in symptomatic post-Fontan patients and addressing them appropriately. The technique used here could be beneficial for other patients with similar anatomical conditions.
CT angiography showing multiple small vessels around both coronary arteries draining into the main pulmonary artery trunk. Blue, green, yellow arrows indicating left coronary artery, right coronary artery and coronary to pulmonary artery fistulas, respectively.

Angiographic images. ( a ) Aortic root angiography showing fair amount of left to right shunt via multiple micro-fistulas from both coronary arteries to the main pulmonary artery. ( b ) Inferior vena cava angiography. Left pulmonary arteries frow is pushed back by flow from fistulas. ( c ) Remnant of the main pulmonary artery trunk. ( d ) Aortic root angiography after two Amplatzer Vascular Plug II was placed. The amount of left to right shunt was significantly reduced. ( e ) Residual left Blalock-Taussig-Thomas shunt was completely closed by Amplatzer Vascular Plug IV. ( f ) Inferior vena cava angiography a year after the procedure showing improved left pulmonary artery flow and completely occluded main pulmonary artery trunk.

Acknowledgements
The authors would like to thank our colleagues who contributed to the care of the patient. We are also grateful to Dr Hidenori Yamamoto for useful discussion and advice.
Financial support
The authors did not receive support from any organisation for the submitted work.
Competing interests
The authors declare no conflict of interest.
Ethical standards
The authors assert that all work reported complies with the ethical standards of the Helsinki convention, and institutional and national research committee.
