Coronary-to-Pulmonary Artery Fistula

To the Editor:

We read with interest the article by Drs. Wong, Chua, and Chan titled Coronary-Pulmonary Arteriovenous Fistula Used as Proximal Anastomotic Site for Saphenous Vein Grafts in Patient with Porcelain Aorta. ¹ We too have encountered patients with various coronary anomalies, most recently a 53-year-old man who presented with progressive symptoms of angina, shortness of breath, diaphoresis, and fatigue.

Electrocardiography suggested nonspecific ST segment changes in the anterior leads. Coronary angiography revealed significant triple-vessel disease with well-preserved left ventricular function (ejection fraction, 0.70). Selective angiography of the right coronary artery (RCA) revealed a small but codominant vessel exhibiting an 80% stenosis at its mid-portion, as well as a 60% narrowing at the crux. Selective angiography of the left coronary system demonstrated a fistula arising from the left anterior descending (LAD) coronary artery just proximal to a 50% to 60% lesion. The fistula appeared to empty into the main pulmonary artery (Fig. 1). The circumflex artery was narrowed by 60% proximally, and by 90% more distally near the takeoff of the 2nd obtuse marginal branch. Right heart catheterization revealed normal pulmonary artery pressures (30/16 mmHg), with no step-up in oxygen saturations from right ventricle to pulmonary artery (70.5% vs 72.4%). As a consequence, we postulated that the fistula, although small, produced a “steal” phenomenon that, in addition to the underlying coronary artery disease, contributed to the patient's anterior wall ischemia and chest pain.

Fig. 1 Selective angiography shows a fistula (arrow) arising from the left anterior descending coronary artery and appearing to empty into the main pulmonary artery.

At operation, 3 bypass grafts were constructed: left internal mammary artery to the LAD, reverse saphenous vein to the distal RCA, and reverse saphenous vein to the obtuse marginal branch. Conventional cardiopulmonary bypass was used, including moderate systemic hypothermia, cold blood cardioplegia, and the single aortic cross-clamp technique. Attention was then turned to the coronary fistula, which appeared to be entirely epicardial in nature. At its origin, the fistula arose from the LAD as a superfluous mesh of cirsoid vessels. However, this angiomatous network appeared to converge into a single lumen as it entered the main pulmonary artery. Single 4-0 polypropylene sutures with pledgets were placed around the fistula, proximally and distally. In addition, the fistulous tract was obliterated lengthwise with a running 5-0 polypropylene suture.

Simple ligation of a fistula between a coronary artery and a pulmonary artery was 1st described in 1947 by Biorck and Crafoord. ² This surgical approach is satisfactory, provided that the fistula is single lumen in nature. If multiple lumens are suspected, or if the sites of origin and termination are not clearly defined, the pulmonary artery should be opened and the fistula closed under direct vision using cardiopulmonary bypass. ^3–5 According to de Nef and colleagues, ⁶ the majority of coronary fistulas that arise from the native RCA empty into the right atrium or right ventricle. Wong, Chua, and Chan's description of a RCA-to-pulmonary artery fistula is a rare find indeed. ¹ Furthermore, we applaud their creative use of this anomalous vessel as a proximal anastomotic site in a patient with a porcelain aorta.