Coronary artery fistula (CAF) is a rare congenital cardiac anomaly in which a communication is present between a coronary artery and either a cardiac chamber or another vascular structure. It accounts for 0.2–0.45% of all congenital cardiac anomalies.1
Acquired fistulae are rare and can result from thoracic trauma and postinterventional procedures such as endomyocardial biopsy or coronary angioplasty.1,2 It can be complicated by aneurysmal dilatation and atherosclerosis of the affected vessel.1 Surgical ligation of the fistula is considered the gold standard; however, percutaneous catheter closure has become an accepted alternative to surgery. Closure of the fistula is safe and effective; however, it can be associated with significant morbidity and mortality.
Case Report. A 48-year-old female patient was referred to our institution for transcatheter closure of a coronary artery fistula after an unsuccessful attempt at another institution. The fistula was diagnosed after she presented to the other institution with shortness of breath and fatigue; a circumflex coronary artery-to-coronary sinus fistula was diagnosed by transesophageal echocardiography (TEE) and confirmed by angiography. At that time, she had normal left ventricular (LV) wall motion with an estimated ejection fraction (EF) of 65%. The patient’s past medical history was remarkable for mitral valve prolapse, but no regurgitation. She was on an oral contraceptive. Her cardiac examination was remarkable for the presence of a 2/6 continuous murmur heard best at the lower sternal border; her electrocardiogram (ECG) was normal.
The patient underwent her second cardiac catheterization at our institution. Access was obtained from the right femoral vein, the right femoral artery and the right internal jugular vein. Her right and left heart pressures were normal, and the calculated Qp/Qs ratio was 1.1:1. Coronary angiography demonstrated a dilated left main coronary artery (LMCA) and dilated circumflex coronary artery measuring 6 mm, with a fistula ending in the coronary sinus just before it joined the right atrium (Figure 1). Using a retrograde approach and under fluoroscopic as well as intracardiac echocardiographic guidance (Figure2), the fistula was embolized using a total of 7 coils: 1 Nester coil (8 mm x 14 cm), 2 Nester coils (6 mm x 14 cm), 2 Gianturco coils (6 mm x 4 cm), and 2 Gianturco coils (3 mm x 3 cm) (Cook, Inc., Bloomington, Indiana). Postembolization angiography demonstrated complete occlusion of the fistula (Figure 3). She was discharged home the next day in good condition, receiving no antiplatelet or anticoagulation therapy.
The patient flew back to her home town 4 days after discharge when she contacted us complaining of chest pain that apparently started the day after discharge. The pain was left-sided and was worse with inspiration and associated with mild shortness of breath. Evaluation at a local hospital revealed that she had an acute myocardial infarction (CK-MB 24.7 ng/ml and troponin I 29.9 ng/ml), with echocardiographic evidence of left ventricular (LV) posterior/inferior wall hypokinesia and moderate mitral valve regurgitation, as well as ECG evidence of atrial fibrillation with a rate of 103/minute. Amiodarone and digoxin therapy were initiated to control the rate, and a heparin infusion was started to prevent thrombus formation. A day-and-a-half later, she was flown to a neighboring state where she underwent cardiac catheterization and cardioversion after TEE evaluation. Coronary angiography revealed complete occlusion of the left circumflex coronary artery to its origin from the left main coronary artery (Figure 4). Since this occlusion and infarct were a few days old, no further attempt was made to recanalize the occluded artery, and she was discharged home in normal sinus rhythm on amiodarone, ACE-inhibitor, beta-blocker, furosemide, aspirin and nitroglycerin.
Two weeks later, the patient presented with right-sided pleuritic chest pain, worse with inspiration, with intermittent episodes of palpitations. Investigations showed markedly elevated D-dimers; computed tomographic angiography of the chest showed evidence of a pulmonary embolism in the right lung, and a small clot was seen on the left side. Ultrasound Doppler of the legs showed no evidence of deep-vein thrombosis. She was started on enoxaparin and warfarin, with the possibility of intermittent episodes of atrial fibrillation.
Discussion. CAF was first described by Krause in 1865.3 Little is known about the natural history of CAF and long-term outcomes.3,4 In a consecutive series of 14,708 coronary angiograms examined for the presence of coronary artery fistulae, the incidence was 0.13%.5
CAF most commonly originates from the left coronary artery (LCA) in 53–67% of cases,5–7 with the LAD being the most common origin, followed by the RCA and circumflex coronary artery in 33%. Rarely, bilateral CAFs originate from both the RCA and LCA in 6.6–16% of cases.7,8 The most common site of drainage is the right side of the heart, with the right ventricle (RV) being the site in 44–50% of cases, and the right atrium (RA) in 20–36%, followed by the left atrium (LA) and pulmonary artery (16%), and rarely into the LV (8%) or CS.6–8 Associated cardiac defects are present in 5–30% of the patients, with atrial-septal defect (ASD) or patent ductus arteriosus (PDA) being the most common lesions.8
Symptomatic patients usually present with congestive heart failure, dyspnea on exertion, angina or myocardial ischemia, which was reported even in pediatric patients,9 and rarely endocarditis and aneurysmal dilatation and rupture.10,11 In recent surveys, 73% of patients were symptomatic, with overall fistula-related complications of 21%, and fistula-related mortality of 6%.6 Older patients are more likely to be symptomatic, with the incidence increasing after the age of 20 years.3 In some studies, invasive testing demonstrated that there was no difference in the Qp/Qs ratio between symptomatic and asymptomatic patients,6,8,10 but older patients were more likely to have elevated pulmonary artery (PA) pressure and LV end-diastolic pressure.
In the presence of CHF, significant left-to-right shunt, or arrhythmias, elective closure of CAF is accepted,11 however, it is still controversial in asymptomatic patients.3,12 In a large literature review in 1979, Liberthson et al reported an operative mortality rate of 1% in patients < 20 years of age as compared to 7% in those who were > 20 years of age; also, operative complications were much higher in the older age group (7% vs. 23%, respectively). They also found that younger patients were less likely to be symptomatic or to have complications related to CAF (19%) as compared to 63% in patients > 20 years of age.10 Due to this and the high incidence of late symptoms and potentially fatal complications,1 as well as the increased morbidity and mortality with ligation in older patients,6,8 some recommended that all patients with CAF should undergo closure of their fistulae.
Surgical closure of CAF is safe and effective,3 with lower mortality and morbidity rates than the natural history of the disease, especially if surgery is performed at a young age.13 The goal is abolition of the fistulous connection without interfering with the native coronary artery supply.6 The optimal surgical procedure and the need for cardiopulmonary bypass (CPB) depend on the fistula’s size, location and drainage pattern, and the presence of associated cardiac lesions. CPB is needed in 56–67% of the patients who undergo surgical closure,2 especially in the presence of multiple entry points, when exposure is technically difficult, or when aneurysmectomy or bypass grafting is needed.6
In the current era of percutaneous interventions, transcatheter embolization is reported to be an effective alternative to surgery.13 Several transcatheter techniques have been described, including: standard steel coils (Cook Inc.); Gianturco coils (Cook Inc.); platinum microcoils (Cook Inc.); detachable balloons; polyvinyl alcohol foam; and double-umbrella devices, including the Amplatzer line of devices.2,7,14–16,19–21 The most commonly used technique/device is coil embolization.2 Certain characteristics of the fistula must be met for transcatheter closure to be successful: (1) the feeding coronary artery branch must be able to be safely cannulated either from the RV cavity or retrogradely from the coronary ostium; (2) the fistula should preferably be single and narrow, with a restrictive draining site; (3) no vital coronary branch originates at the site of potential device deployment; and (4) an absence of multiple fistulous communications.14 Percutaneous closure of CAF was not feasible in 15% of patients according to one study.2 Percutaneous closure is as effective as surgery in closing CAF, with an overall complete closure rate of 82%, and a partial closure rate with a small residual shunt in 18% of patients,2 compared to residual or recurrent shunt in 4–28% after surgical closure.6 This even holds true for small infants and children.16
Complications that have been reported after transcatheter closure include transient arrhythmias and minor ST-T changes in up to one-third of patients,2 and migration of the coil(s) or device to the RV or PA. One patient had a myocardial infarction (MI). A mortality rate of 2.2% has been reported after embolization of the device and dissection of coronary artery.2
It has been shown in different studies that most affected coronary arteries were dilated and tortuous, and postoperative follow up showed that affected coronary arteries remained dilated and tortuous in 90% of patients, with later development of coronary thrombus, calcification and ventricular aneurysm in some patients at long-term follow up.9 Some concluded that ligation of the fistulous termination alone does not reduce the size of the fistula.
Myocardial infarction has been reported rarely in the literature after either surgical or percutaneous closure of CAF. In one surgical series, follow-up catheterization after surgical closure of CAF showed that 19% of patients had evidence of thrombosis of a segment of coronary artery. It was also shown that the dilated proximal coronary arteries remained dilated or were thrombosed all the way to their aortic origin.3 In another report of long-term follow up of 7 patients with small-to-moderate left-to-right shunts through CAF, one patient developed spontaneous occlusion of the RCA proximal to the fistula, but remained asymptomatic.4 Another study reported on 3 patients who underwent surgical closure during childhood,2 of these, 2 patients had a marked decrease in the size of the aneurysmal coronary artery that supplied the fistula after successful closure; however, they abnormal myocardial perfusion at their last follow-up examinations, with intimal thickening documented by intravascular ultrasound.13 There is also 1 case report of spontaneous thrombus formation in an untreated CAF from the RCA to the RA brought to attention by a cerebral infarction.17 There is another case report of a 69-year-old female who developed acute MI several hours after percutaneous closure of a CAF from the RCA to the RA.18
No studies have addressed the risk of spontaneous thrombosis of the feeding coronary artery after either surgical or percutaneous closure of CAF and whether these patients would require long-term antiplatelet or anticoagulant therapy. In view of the progression to coronary obstruction and myocardial ischemia,13 one of the investigators emphasized the importance of coronary angiography and myocardial scintigraphy on long-term follow up of patients after surgery to treat CAF. Others have recommended postoperative antiplatelet therapy, especially in patients with distal CAF and abnormally dilated coronary arteries.9
Our case emphasizes the need, at least, for antiplatelet agents administration after the closure. When this patient was discharged home the following day, our routine at that time was not to administer antiplatelet agents for the fear of fistula recanalization. However, she started to complain of chest pain that was brought to our attention 4 days later, and by the time further evaluation was done, it was too late to salvage the artery and myocardium. Therefore, our practice now has changed to recommend that in the presence of a significantly dilated coronary artery (double the size of the normal coronary artery), the patient should be hospitalized for at least 48 hours, with daily echocardiograms to assess the myocardial function and perform frequent checks of cardiac enzymes. Also, such patients should receive heparin overnight and when discharged, possibly warfarin for several months, and aspirin indefinitely. Whether this strategy of anticoagulation and antiplatelet agents will change the natural history after closure remains to be seen.
- Choi B, Chang H, Choi S, et al. A coronary artery fistula with saccular aneurysm mimicking a right atrial cystic mass. Japan Heart J 2004;45:697–702.
- Armsby LR, Keane JF, Sherwood MC, et al. Management of coronary artery fistulae: Patient selection and results of transcatheter closure. J Am Coll Cardiol 2002;39:1026–1032.
- Cheung D, Au W, Cheung H, et al. Coronary artery fistulas: Long-term results of surgical correction. Ann Thorac Surg 2001;71:190–195.
- Jaffe R, Glancy DL, Epstein S, et al. Coronary arterial-right heart fistulae: Long-term observations in seven patients. Circulation 1973;47:133–143.
- Gillebert C, Van Hoof R, Van De Werf F, et al. Coronary artery fistula in an adult population. Eur Heart J 1986;7:437–443.
- Hong GJ, Lin CY, Lee CY, et al. Congenital coronary artery fistulas: Clinical considerations and surgical treatment. Ann J Surg 2003;74:350–355.
- Trehan V, Yusuf J, Mukhopadhyay S, et al. Transcatheter closure of coronary artery fistulas. Indian Heart J 2004;56:132–139.
- Shyam Sunder KR, Balakrishnan KG, Tharakan JA, et al. Coronary artery fistula in children and adults: A review of 25 cases with long-term observations. Int J Cardiol 1997;58:47–53.
- Wang NK, Hsieh LY, Shen CT, et al. Coronary artery fistula in pediatric patients: 17-year institutional experience. J Formos Med Assoc 2002;101:177–182.
- Liberthson R, Sagar K, Berkoben J, et al. Congenital coronary arteriovenous fistula: Report of 13 patients, review of the literature and delineation of management. Circulation 1979;59:849–854.
- Pedra CA, Pihkala J, Nykanen DG, et al. Antegrade transcatheter closure of coronary artery fistulae using vascular occlusion devices. Heart 2000;83:94–96.
- Carrel T, Tkebuchava T, Jenni R, et al. Congenital coronary artery fistulas in children and adults: Diagnosis, surgical technique and results. Cardiology 1996;87:325–330.
- Hiraishi S, Misawa H, Horiguchi Y, et al. Effect of suture closure of coronary artery fistula on aneurysmal coronary artery and myocardial ischemia. Am J Cardiol 1998;81:1263–1267.
- Eicken A, Sebening W, Genz T, et al. Coil embolization of congenital coronary arterial fistulas. Pediatric Cardiol 2003;24:149–153.
- Kambara AM, Pedra CA, Esteves CA, et al. Transcatheter embolization of congenital coronary arterial fistulas in adults. Cardiol Young 1999;9:371–376.
- Holzer R, Waller BR, Kahana M, et al. Percutaneous closure of a giant coronary arteriovenous fistula using multiple devices in a 12-day-old neonate. Catheter Cardiovasc Interv 2003;60:291–294.
- Preib M, Habicht J, Bongartz G, et al. Aneurysmal and partially thrombosed orifice of a coronary artery fistula into the right atrium combined with patent foramen ovale. Thorac Cardiovasc Surg 2001;49:120–121.
- Ascoop AK, Budts W. Percutaneous closure of a congenital coronary artery fistula complicated by an acute myocardial infarction. Acta Cardiologa 2004;59:67–69.
- Al-Ata J, Amin M, Galal MO, et al. Transcatheter occlusion of a large left coronary artery to right superior vena cava using the Amplatzer duct occluder device. Pediatr Cardiol 2004;25:70–72.
- Thomson L, Webster M, Wilson N. Transcatheter closure of a large coronary artery fistula with the Amplatzer duct occluder. Catheter Cardiovasc Interv 1999;48:188–190.
- Khan MD, Qureshi SA, Rosenthal E, et al. Neonatal transcatheter occlusion of a large coronary artery fistula with Amplatzer duct occluder. Catheter Cardiovasc Interv 2003;60:282–286.