Discussion:
When a fistula drains into the right side of the heart, the volume overload is occurring there, as well as in the pulmonary vascular bed, left atrium, and left ventricle. When the fistula drains into the left atrium or ventricle, volume overloading takes place in these chambers, but pulmonary blood flow does not dramatically increase. The size of the shunt is determined by the size of the fistula and the pressure difference between the coronary artery and the chamber into which the fistula drains. Most patients display no symptoms, especially during the first two decades of life, and are referred for testing due to an asymptomatic murmur; however, older patients may experience dyspnea or angina and occasionally arrhythmias. Complications that may arise include ‘theft’ from the adjacent myocardium, which can lead to myocardial infarction (MI), thrombosis and embolism, heart failure, atrial fibrillation, rupture, endocarditis/endarteritis, and even sudden death [4,5]. When diagnosing an AV shunt, CXR and EKG do not generally have significant value [6]. When the initial EKG is normal, treadmill or nuclear stress tests often reveal ischemic changes. 2D Doppler echocardiography and TEE may also show the site of drainage, but it is difficult to understand the detailed anatomy using these techniques. [4,5,6] Cardiac MRA is limited because the information regarding the origin and drainage site of the fistula are less clear in comparison with those observed with CTA and conventional angiography [7]. Previously, cardiac catheterization was the first-line modality for diagnosing CAFs and was thought to have the highest efficiency for detailing the anatomy, size, origin, course, and presence of any stenosis and/or drainage; however, more recent studies have shown that CTA actually produces the best results. It has been demonstrated that the exact 3D course of the artery is difficult to obtain by conventional angiography, the drainage sites may not be as well visualized, and the invasive nature yields a higher mortality rate than CTA [7]. CTA is a relatively new imaging tool that was introduced 20 years ago for non-invasive cardiac imaging, and with the introduction of multi-detector computed tomography (MDCT), many problems related to image quality have been resolved [8]. CTA can produce high quality images that clearly delineate the cardiac chambers and coronary arteries and veins [5], in addition to having a shorter acquisition time and higher spatiotemporal resolution. The multiplanar reformation clearly demonstrates the sites of origin and termination of abnormal blood vessels, and 3D data provide a more precise overview of the heart, its vascular anatomy, and information pertaining to the diagnosis of coexistent abnormalities as compared with left heart catheterization [7]. However, limitations include less hemodynamic information, which we speculate to be the rationale for its failure to detect the fistula in our case.
According to the 2008 ACC/AHA guidelines, all large fistulas and small-to-moderate fistulas associated with myocardial ischemia, systolic/diastolic dysfunction, or ventricular dilation should undergo closure irrespective of the symptoms of lack thereof [9]. Patients diagnosed with CAFs should be educated prior to treatment. In contrast to patients with acquired heart disease, those with CAFs or adult congenital heart disease (ACHD) may have never experienced ”normal” function. Cardiopulmonary exercise testing provides an objective assessment and has been shown to possess prognostic value for a wide variety of congenital conditions. There exist conflicting arguments regarding physical activity in these patients, with some suggesting restricted exercise, which will inevitably lead to a greater prevalence of obesity and other forms of heart disease. The most recent guidelines recommend safe exercise in patients across the spectrum of CAF severity, being individualized depending on patient clinical status. Self-directed activity is usually at 40−60% of the maximal exercise capacity, whereas fitness training occurs at 60−80% [9]. The treatment options for CAFs include surgery or catheter closure [5]. The main surgical options are epicardial fistula mobilization with ligation or division, arteriotomy of the dilated proximal coronary artery with suturing close to the fistula origin, and exposure of the fistulous connection and closure by direct suture or autologous pericardial patch. However, surgery is associated with a morbidity and mortality rate ranging from 0−6%, and there is a risk of MI and recurrence of the fistula [10]. Catheter closure is considered an effective and safe alternative, with the aim of occluding the fistula artery as distally and close to its termination point as possible to avoid occlusion of the branches supplying the normal myocardium [10]. There are many tools available for closure, including occlusion coils, detachable balloons, and deployable stents and balloons. The techniques and tools used for catheter closure are influenced by several different factors such as the morphology of the feeding arteries, patient age, and location of the fistula [10]. With catheter closure techniques, complete occlusion of the fistula may be achieved in >95% of patients. In the remaining patients, either further procedures are required to close the fistulas or conservative management can be undertaken if the residual fistulas are small [11]. Selection of percutaneous or surgical management is influenced by the anatomy and physiology of the CAF. Indications for surgical management are large, symptomatic, and/or tortuous fistulas with or without multiple communications and drainage; whereas, percutaneous transcatheter closure is considered when the origin of the fistula is in the proximal portion of the feeding coronary artery, is non-tortuous, or is in an older patient with high preoperative risk [7]. After closure, these patients should be placed on Coumadin, with a goal therapeutic INR of 2.5 for 6−12 months; and should be on anti-platelet drugs indefinitely.
Conclusions: The present case illustrated a rare CAF, in which the drainage is into the left circulation. CAFs are only detected in 0.05−0.25% of patients who undergo coronary angiography and left-sided drainage is seen in a mere 10% of these cases. Although current literature states that CTA is the first-line modality for the detection of CAFs and provides a more precise overview of the heart, its vascular anatomy and information pertaining to coexisting abnormalities via 3D data, we found that conventional angiography was able to detect the fistula when CTA was not. Further research should be elicited to directly compare the sensitivity and specificity of coronary angiography with that of CTA for the detection of CAFs.