Introduction
Congenital fistulae cover a wide spectrum of vascular disorders including a variety of arteriovenous (AV) malformations, but also can include venovenous communications. Structural and congenital cardiology interventionalists most often may be asked to evaluated those fistulae involving the heart, great vessels, and pulmonary circulation. Because of the widespread nature of such vascular malformations, it is difficult to determine an exact epidemiologic incidence of such disorders, although they may be found with significant frequency during routine computed tomography (CT) screening for other disorders. Perhaps the most frequent clinically significant malformations involve the coronary and pulmonary circulation. The basic hemodynamic effect observed with such malformations is an increase in cardiac output resulting in both right and left ventricular volume overload.
Fundamentally, flow will occur whenever there is a pressure gradient between two interlinked structures. Pressure differences between the heart’s chambers are dynamic and may vary with volume state, pulmonary or systemic resistance, and throughout the cardiac cycle. Moreover, the heart may change in relation to a shunt over time. Longitudinal assessment over time may be necessary to realize the impact of a defect.
The most common fistula encountered often in routine coronary angiography for coronary artery disease assessment is a tiny coronary fistula to the front of the main pulmonary artery. These fistulae likely represent a failure of involution or persistence of coronary artery anlagen in the pulmonary artery, similar to that observed with anomalous origin of the coronary artery from the pulmonary artery. There is a reported incidence of 0.2% to 0.4% of congenital cardiac defects. The tiny fistula from the left or right coronary artery to the main pulmonary artery is generally hemodynamically insignificant and often has multiple sources, such as from both coronary arteries or brachiocephalic branches, as seen in Fig. 27.1 . These tiny fistulae are difficult to embolize but technically can be coiled if necessary. In general, they are asymptomatic and do not require intervention.
Coronary artery fistulae are rare abnormal communications between a coronary artery and a coronary chamber (coronary-cameral fistula) or a major vessel (coronary-AV fistula). Most commonly these are congenital in nature; however, there is a gradual increase in acquired coronary fistulae resulting from infections, trauma, or iatrogenic causes. They often are found with very large dilated and tortuous proximal coronary arteries, as shown in Fig. 27.2 . These patients nevertheless often remain asymptomatic and are only recognized later in adulthood.
The architecture of the coronary artery is affected by the location of the fistula. If the fistula is proximal with the coronary artery, it results in dilatation of the proximal artery with a relatively normal distal artery. A distal fistula typically results in dilatation of the entire artery.
Potential complications of coronary fistulae include massive coronary artery dilatation and aneurysm formation due to the continuous high-flow, intracoronary thrombus, which may result in coronary occlusion and myocardial infarction, coronary rupture, arrhythmia, and infective endocarditis (due to AV shunting). Patients with a coronary-to-bronchial artery fistula may present with hemoptysis or bronchiectasis.
Most commonly a coronary artery fistula will have a single origin and a single outflow; however, complex fistulae may have multiple feeding arteries or multiple drainage sites. The most common drainage sites are the pulmonary trunk/pulmonary arteries, right ventricle, or right atrium. Spontaneous closure is rare but has been reported.
This chapter will present a standardized approach to these defects in four sections: (1) clinical assessment and decision to close; (2) preprocedure assessment, defining the defect, and procedural planning; (3) procedural techniques; and (4) clinical outcomes and follow-up assessment.
Clinical assessment
Most often, given the sensitivity and availability of modern imaging, the structural interventional cardiologist is presented with a discovered defect and posed with two questions: (1) Can this defect be closed? and (2) Should this defect be closed? In this section we will look critically at the decision to intervene.
Because this group of defects is so diverse, the operator must individualize the balance of risk and benefit for each unique situation. To accurately assess the risk and potential benefits of defect closure, one must evaluate the impact of the defect on the patient’s heart function in the context of adjacent cardiac disease and the natural history if left untreated. For example, a relatively small shunt may contribute to substantial symptoms in a patient with diastolic disfunction, whereas the same-size shunt in another patient may be well tolerated without the need for intervention.
In principle, patients with defects that cause exercise limitations, change in chamber function (or enlargement), or heart failure are likely to benefit from successful closure. The clinical assessment should focus on the presence of these symptoms and findings. Auscultation is less reliable for these defects because even high-flow fistulas may be inaudible.
Although no single laboratory value will dictate the decision to intervene, the additional information is always valuable. Laboratory assessments should include assessment of changes in renal function, liver function, hemoglobin, and heart failure biomarkers (e.g., B-type natriuretic peptide). Anemia may suggest acquired von Willebrand disease attributable to high defects causing high shear stress on the blood, which can be confirmed by testing the distribution of von Willebrand multimers.
In the asymptomatic patient without heart failure or change in heart function, an exercise study is a useful tool to quantify exercise capacity. Simultaneous measurement of VO 2 can increase the precision and value of exercise testing.
Invasive hemodynamic testing allows objective assessment of left and right heart filling pressures and pulmonary and systemic resistance and to calculate shunt fraction. Although no single shunt value demands repair or suggests deferment, in principle, the larger the shunt flow, the greater the physiologic implication and likelihood of benefit with closure. Qp/Qs ratios greater than 1.5 to 2.0 have been used historically with decision-making in the setting of atrial septal defects and represent a starting point when incorporating into decisions about whether to proceed with closure.
AHA guidelines
The 2008 American Heart Association (AHA) guidelines for the management of adults with congenital heart disease are provided in the following box. The 2018 guidelines were less detailed and stated that the presence of coronary artery fistula(s) requires review by a knowledgeable team that may include congenital or noncongenital cardiologists and surgeons to determine the role of medical therapy and/or percutaneous or surgical closure.
| A large coronary arteriovenous fistula (CAVF), regardless of symptomatology, should be closed via either a transcatheter or surgical route after delineation of its course and its potential to fully obliterate the fistula | I | C |
| A small-to-moderate CAVF in the presence of documented myocardial ischemia, arrhythmia, otherwise unexplained ventricular systolic or diastolic dysfunction or enlargement, or endarteritis should be closed via either a transcatheter or surgical approach after delineation of its course and the potential to fully obliterate the fistula | I | C |
| Clinical follow-up with echocardiography every 3 to 5 years can be useful for patients with small, asymptomatic CAVF to exclude development of symptoms or arrhythmias or progression of size or chamber enlargement that might alter management | IIa | C |
| Patients with small, asymptomatic CAVF should not undergo closure | III: Harm | C |
Procedural planning
Clearly defining the size, location, and extent of a given defect is vital for procedural planning. Modern imaging techniques have become so clear that if a defect cannot be clearly characterized with a combination of CT and transesophageal echocardiography (TEE), then the likelihood of successful closure is small. Imaging analysis should focus on (1) defining the defect anatomy, (2) adjacent anatomy (e.g., myocardial coronary branches, valve apparatus, conduction system, etc.), and (3) physiologic impact of the defect (i.e., chamber size).
Before deciding on intervention, it is important to establish whether the fistula has a single/multiple inflow and/or a single/multiple outflow. Imaging will assess the degree of tortuosity and length of catheter likely to be needed if percutaneous closure is performed.
Transthoracic echocardiography (TTE) is widely available and relatively inexpensive. More posterior or basal structures can be challenging to characterize, and acoustic shadowing (from calcium or prosthetic material) may compromise some imaging. Most patients have had a TTE before or at diagnosis; however, for those who have not, it is advised to include preprocedure TTE as a baseline.
Transesophageal echocardiography (TEE) is invasive but particularly valuable at defining intracardiac defects and is often useful for guiding procedural closure. TEE provides high-resolution images of posterior cardiac structures along with Doppler-based flow measurements. More apical structures are in the far field and may be better imaged with TTE. Three-dimensional TEE imaging can be particularly useful for identifying the relationship of a defect to adjacent anatomy or prosthetic material. Moreover, integration of TEE color flow imaging with cross-sectional imaging provides the operator with a clear sense of defect size and drainage location.
CT: The combination of spatial resolution with cross-sectional imaging provides three-dimensional visualization of anatomy. With an electrocardiogram (ECG)-gated CT data set it is possible to thoroughly assess the size, extent, and location of a defect, noting entry and exit sites and surrounding structures. One can simulate the fluoroscopic angle to find the optimal view for use during implant and the relation to adjacent fluoroscopic markers (i.e., adjacent prosthetic valves or leads).
Multiple rotational views of the entire fistula, the associated coronary artery branches, and the final communications can be obtained to allow a better understanding of the complexity and course of the fistula. Fig. 27.3 and Video 27.1 illustrate a coronary CT angiogram in a complex right coronary artery (RCA) fistula to the coronary sinus (CS). There are multiple twists and turns in the dilated RCA, and several terminal right ventricular branches are present distally in the coronary. There also is a complex entry of the RCA fistula into the CS.
