Large Spontaneous Coronary Artery-to-Right Ventricular Fistula

Coronary fistulae are congenital or acquired communications between a coronary artery and either a chamber of the heart (coronary-cameral fistula) or any segment of the systemic or pulmonary circulation bypassing the myocardial capillary network. Coronary artery fistulae are a rare occurrence seen in only 0.1–0.2% of angiograms, but comprise about 14% of all congenital coronary anomalies.^1–3 Most fistulae are congenital in origin,^3–5 but acquired fistulae have been rarely reported as a consequence of trauma,^4,6 coronary artery bypass surgery,⁷ complication of percutaneous coronary intervention^8,9 and myocardial infarction.^10,11 These fistulae may be solitary or multiple, originating from one or more coronary arteries.¹ Spontaneous fistulae are extremely rare.^12,13 We report a case of the spontaneous formation of a left anterior descending artery (LAD) to right ventricular (RV) fistula without any precipitating event. This spontaneous occurrence was documented on serial angiography and confirmed during surgery. Case Report. In September of 1995, a 68-year-old African-American male presented to his primary cardiologist for preoperative evaluation prior to a L4-L5 laminectomy procedure. An exercise thallium stress test was performed revealing ischemia in the inferior and lateral distributions. The patient was referred for coronary angiography to evaluate his coronary anatomy. The catheterization showed mild LAD disease with 90% stenosis of the first diagonal artery. The left circumflex artery (LCx) had a subtotal occlusion in the proximal portion with TIMI-1 flow in the distal vessel. The right coronary artery (RCA) had diffuse irregularities with 60% narrowing of the posterior descending artery (PDA). Medical management was recommended and he underwent surgery without complications. In April 1998, the patient presented to a local emergency department with complaints of progressive exertional angina for 2 to 3 days. On the day of presentation, his pain was unrelenting and associated with dyspnea and diaphoresis. He was transferred to our facility for further treatment. He underwent cardiac catheterization, which identified a normal left main artery. The proximal and mid LAD had moderate, diffuse disease. The first diagonal branch was subtotally occluded with distal filling occurring via right-to-left collaterals (Figure 1). The LCx was unchanged from the previous catheterization. The RCA was diffusely diseased with 70–80% narrowing of the mid and distal vessel. The mid-PDA had 70–80% stenosis. Left ventriculography showed anterolateral and inferobasal hypokinesis with an estimated ejection fraction of 40–45%. The patient’s mean right atrial (RA) pressure was 10 mmHg, his RV pressure was 29/12 mmHg, his pulmonary artery pressure was 29/10 mmHg and his pulmonary capillary wedge pressure (PCWP) was 9 mmHg. The RA, pulmonary artery and aortic oxygen saturations were 77%, 75% and 96%, respectively. Cardiothoracic surgery was consulted for coronary artery bypass surgery. On the night of cardiac catheterization, the patient developed severe substernal chest pain with changes in the inferior leads on electrocardiography. He underwent emergency cardiac catheterization, resulting in angioplasty of the distal RCA. Surgery was deferred. Three weeks later, he was readmitted for increasing chest pain. Repeat angiography revealed no new changes in the LAD circulation (Figure 2), continued occlusion of the first diagonal and LCx and a patent RCA. Consequently, medical therapy was continued with an emphasis on risk factor modifications. In September 2002, the patient underwent catheterization after presenting with worsening angina and shortness of breath. The LCx was unchanged. The LAD had a 60–70% complex lesion in the proximal portion. The first diagonal remained occluded. The patient’s RCA was diffusely diseased, with a 70-80% lesion *at the distal bifurcation. A large fistulous connection between the distal LAD and the RV was noted (Figure 3). This fistula was not present on the patient’s prior catheterizations. Right heart catheterization was not performed at this time. He was referred for coronary artery bypass surgery. Saphenous vein grafts were placed to the PDA and LCx. The left internal mammary artery (LIMA) was anastomosed to the LAD. The endocardial surface of the RV was inspected without identification of an apparent fistula. The clamp on the LIMA was removed, allowing visualization of blood flowing from the LAD to the septal surface of the mid-RV just beneath the moderator band (Figures 4 and 5). The area was ligated with a pledgeted 4–0 proline suture in a horizontal mattress fashion. After suture placement, very little blood flow was noted in the RV with injection of cardioplegia through the LIMA (Figure 6). The patient tolerated the procedure without difficulty, and the remainder of his hospital course was uneventful. In October of 2005, he was referred for repeat cardiac catheterization secondary to recurrence of his angina symptoms. A right heart catheterization was performed. His mean right atrial (RA) pressure was 10 mmHg, RV pressure was 30/10 mmHg, pulmonary artery pressure was 30/12 mmHg and PCWP was 12 mmHg. The RA, RV, pulmonary artery and arterial oxygen saturations were 67%, 68%, 69% and 99%, respectively. The angiographic portion of the study showed patency of all grafts. Competitive flow was noted in the distal portion of the LAD. A fistulous connection between the LAD and RV was again noted upon injection of the native LAD (Figure 7) and LIMA (Figure 8), but to a lesser degree than was previously noted. A transthoracic echocardiogram was obtained post-catheterization. The LAD-to-RV fistula was appreciated at the apex as diastolic turbulent flow through the interventricular septum (Figure 9). The patient was discharged home without further intervention. Discussion. Spontaneous coronary artery fistulae are an extremely rare occurrence.^12–14 Coronary fistulae are usually single, but cases of multiple fistulae have also been described.¹ The development of coronary fistulae has been described after myocardial infarction^10,11 and in association with left ventricular thrombus.^15,16 It has been hypothesized that these fistulae are the product of neovascularization triggered by the presence of ventricular thrombus producing a vascular channel.¹⁶ Spontaneous fistulae have also been reported as a consequence of trauma,^4,6 coronary artery bypass surgery⁷ and complication of percutaneous coronary intervention.^8,9 Our case demonstrates the appearance of a large spontaneous fistulous connection between the LAD and RV cavity. The spontaneous nature of this fistula was documented on serial angiograms with no interval history of a precipitating event. The exact etiology of the fistula is unclear. It is highly unlikely that the fistula was congenital in nature as the patient underwent multiple angiograms that did not demonstrate any fistulous connection prior to his study in 2002. There was no evidence of thrombus to suggest that this could be the etiology of the fistula. There were no intervening cardiac procedures since the last angiogram, which did not show any fistulous connections. Although right heart catheterization was not performed when the fistula was first discovered, there was no history of congestive heart failure and no signs of hemodynamic changes that could have unmasked a previously dormant fistula. This was a theory that we had proposed as an underlying mechanism in a previously reported case.¹³ It could be hypothesized that in view of the extensive degree, duration of coronary artery disease and progression of disease in the RCA, the microfistulae may be a result of “misguided” collateral circulation from the LAD to the RCA.^17,18 It may also be hypothesized that the fistulous connections between that LAD and the RV may be secondary to circulation via Thebesian veins.^17,18 Thebesian veins are small veins that exist in the walls of all cardiac chambers, but have been found to be in greater number in the RV. They originate within the myocardium and pass through the endocardium, forming a direct connection to cardiac chambers. Coronary-ventricular connections have been described in animal models as well as human studies.^19,20 Therefore, it is possible that these coronary microfistulae are the result of many years of slowly progressing coronary disease. Given the need for collateral circulation, these vessels were recruited to help supply the myocardium with oxygenated blood. This mechanism is similar in theory to the practice of transmyocardial laser revascularization.^17,21 Most fistulae are discovered accidentally and there is rarely evidence of right-to-left shunt on oximetry.^2,3 Although most tend to follow a benign course, potential complications including heart failure and endocarditis may occur. Ischemia secondary to coronary steal has also been described.^11,22 The management of these fistulae, especially when asymptomatic, remains controversial. SBE prophylaxis is usually recommended. Although some advocate closure of fistulae, even in asymptomatic patients,^23–25 most agree that surgical or invasive treatment should target symptomatic patients only. Such treatment may include coil embolization or surgical ligation. Coronary fistulae that have plexiform, complex anatomy may present a challenge to both the surgeon and the interventionalist. In our case, the main indication for surgery was coronary artery disease and revascularization. Given the significant angiographic appearance of the shunt, the fact that it was newly acquired and the potential for complications including steal phenomenon with residual ischemia post-bypass, it was decided to attempt closure of this complex fistula during the surgery. In conclusion, we report a very rare case that clearly demonstrates the spontaneous formation of a large LAD-to-RV fistula. Some previous reports of spontaneous fistulae have “assumed” the spontaneous nature of the fistulae when they were encountered on the first angiogram. In this case, the spontaneous appearance is documented because of serial angiograms. Although the fistula could be a result of extensive coronary artery disease allowing for the development of collateral circulation via the Thebesian venous system or “misguided” collateral circulation, a definitive explanation cannot be offered.

References 1. Yamanaka O, Hobbs RE. Coronary artery anomalies in 126,595 patients undergoing coronary arteriography. Cathet Cardiovasc Diagn 1990;21:28–40. 2. Hobbs RE, Millit HD, Raghavan PV, et al. Coronary artery fistulae: A 10-year review. Cleve Clin Q 1982;49:191–197. 3. Vavuranakis M, Bush CA, Boudoulas H. Coronary artery fistulas in adults: Incidence, angiographic characteristics, natural history. Cathet Cardiovasc Diagn 1995;35:116–120. 4. Gillebert C, Van Hoof R, Van de Werf F, et al. Coronary artery fistulas in an adult population. Eur Heart J 1986;7:437–443. 5. Said SA, el Gamal MI, van der Werf T. Coronary arteriovenous fistulas: Collective review and management of six new cases — Changing etiology, presentation, and treatment strategy. Clin Cardiol 1997;20:748–752. 6. Martin R, Mitchell A, Dhalla N. Late pericardial tamponade and coronary arteriovenous fistula after trauma. Br Heart J 1986;55:216–218. 7. Tami LF. Coronary artery-right ventricular fistula after coronary artery bypass grafting. Clin Cardiol 1993;16:155–157. 8. el-Omar MM, Hargreaves MR, Venkataraman A, Been M. Coronary ventricular fistula as a complication of PTCA: A case report and literature review. Int J Cardiol 1995;51:113–116. 9. Marques KM, De Cock CC, Bronzwaer JG, Visser CA. LAD-right ventricular fistula complicating PTCA: Another case. Cathet Cardiovasc Diagn 1997;42:34–36. 10. Ryan C, Gertz EW. Fistula from coronary arteries to left ventricle after myocardial infarction. Br Heart J 1977;39:1147–1149. 11. Kiuchi K, Nejima J, Kikuchi A, et al. Left coronary artery-left ventricular fistula with acute myocardial infarction, representing the coronary steal phenomenon: A case report. J Cardiol 1999;34:279–284. 12. Ray SG, Cowan MD, Kennedy JA. Spontaneously acquired fistula from the right coronary artery to the right ventricular cavity. Br Heart J 1992;67:323–324. 13. Shammas RL, Miller MJ, Maxwell LP, Babb JD. Spontaneous left anterior descending artery to right ventricular fistula. J Invasive Cardiol 2001;13:467–470. 14. Calick A, Dorsey TJ. Neovascularity related to a mural thrombus. Cathet Cardiovasc Diagn 1979;5:175–177. 15. Nawa S, Soga H, Urakami A, et al. Coronary fistulous communications with left ventricular chamber and giant mural thrombus after myocardial infarction. Jpn Heart J 1994;35:801–808. 16. Kwan T, Feit A. Late neovascularization of longstanding left ventricular mural thrombus — A case report. Angiology 1993;44:647–650. 17. Ansari A. Anatomy and clinical significance of ventricular Thebesian veins. Clin Anat 2001;14:102–110. 18. Taylor JR, Taylor AJ. Thebesian sinusoids: forgotten collaterals to papillary muscles. Can J Cardiol 2000;16:1391–1397. 19. Maric I, Bobinac D, Ostojic L, et al. Tributaries of the human and canine coronary sinus. Acta Anat (Basel) 1996;156:61–69. 20. Esperanca Pina JA, Correia M, O'Neill JG. Morphological study on the Thebesian veins of the right cavities of the heart in the dog. Acta Anat (Basel) 1975;92:310–320. 21. Moosdorf R, Rybinski L, Hoffken H, et al. [Transmyocardial laser revascularization in stable and unstable angina pectoris]. Herz 1997;22:198–204. 22. Cheng TO. left coronary artery to left ventricular fistula: Demonstration of coronary steal phenomenon. Am Heart J 1982;104:870–872. 23. Cheung D, Au W, Cheung H, et al. Coronary artery fistulas: Long term results of surgical correction. Ann Thorac Surg 2001;71:190–195. 24. Liberthson RR, Sagar K, Berkoben JP, et al. Congenital coronary arteriovenous fistulas: Report of thirteen patients. Review of the literature and delineation of the management. Circulation 1979;59:849–854. 25. Goto Y, Abe T, Sekine S, et al. Surgical treatment of the coronary artery to pulmonary artery fistulas in adults. Cardiology 1998;89:252–256.