Importance of Coronary to Pulmonary Artery Fistulae Incidentally Detected on Echocardiography: Can We Ignore It During Childhood?

Backround: Coronary artery fistula (CAF) is an abnormal precapillary connection between a coronary artery and an adjacent structure. The incidence of CAF in children as estimated from echocardiography is 0.06-0.2%. We aimed to establish the follow up results of clinically silent CAF found incidentally by colour Doppler echocardiography.
Methods: The study included patients with abnormal fistulous flow in the pulmonary artery who had been studied between 2008 and 2020. Patient demographics, clinical findings, indication for echocardiography, electrocardiographic findings at presentation, follow-up times and any fistula progress were obtained from the recorded files.
Results: Among the 78,000 patients who had had an echocardiographic examination, 118 had been found by colour Doppler flow mapping to have a clinically silent CAF. The exit point of CAF was clearly visualized with colour Doppler in all patients. In most patients the CAF was located around the pulmonary valve and the bifurcation; in 76 patients, it was on the aortic side of the main pulmonary artery, 26 patients had fistulous flow through the anterior wall of the main pulmonary artery, two had the right ventricular outflow tract (RVOT) as the exit site, for 12, it was the right pulmonary artery, and for another two, the left pulmonary artery. The 118 patients had ongoing follow-up to mean 41 months. There was no spontaneous resolution of fistula. The patients were asymptomatic without intervention and with ongoing echocardiographic evidence of small CAF at the last follow-up.
Conclusion: We recommend the treatment strategy for paediatric patients should be individualized according to fistula origin, size, chamber enlargement, draining site, age of the patient and cost of imaging during follow up. Copyright:

INTRODUCTION

A fistula in the circulatory system is a clearly ectatic vascular segment exhibiting fistulous flow that connects two vascular territories. Coronary artery fistula (CAF) is an abnormal precapillary connection between a coronary artery and an adjacent structure. The abnormal connections can occur between the coronary arteries and cardiac chambers, referred to as “coronary cameral fistulae,” or between the coronary arteries and other vessels, referred to as “coronary arteriovenous malformations.” Typically, a large pressure difference develops between the two territories (the fistula/malformation and the chamber/artery/vessel).[12] Low fistulous flows in the pulmonary artery are usually detected incidentally during transthoracic echocardiography performed for other reasons. While small CAFs are generally diagnosed, color Doppler echocardiography has been reported as a particularly sensitive method to identify these small fistulae.[34]

The incidence of CAF in children as reported in echocardiographic series is estimated at 0.06%–0.2%.[5] The outcome of clinically silent CAFs found incidentally by color Doppler echocardiography, however, is unclear.[6] In addition to individual health issues, the asymptomatic fistulae mentioned in the echocardiography report and followed up by echocardiography in a tertiary center carry psychological and economic burdens. In this review study, therefore, we aimed to establish follow-up results and cost-effectivity for clinically silent CAF found incidentally by color Doppler echocardiography.

MATERIALS AND METHODS

This study included patients with abnormal fistulous flow in the pulmonary artery who had been investigated in outpatient clinics at the Ministry of Health University of Health Science Kanuni Sultan Süleyman Research and Training Hospital and at private practice standard deviation (SD) between January 2008 and February 2020. Full patient data were obtained from the files, covering demographics, symptoms, clinical findings, indications for echocardiography, associated diagnoses, radiologic findings, and electrocardiographic findings at presentation, follow-up times, and progress of the fistula where observed.

Transthoracic echocardiography had been performed with different devices, according to the outpatient clinic involved. The echocardiography machines were equipped with appropriate transducers interfaced with a GE Vivid 3 and 7 ultrasonography system (GE Healthcare, Wauwatosa, Wisconsin, USA) and a Philips IE 33 US system (Philips Healthcare, Andover, Massachusetts, USA). All echocardiographic evaluations were performed in accordance with the American Society of Echocardiography guidelines.[7]

Echocardiography

The pulmonary fistulae had been recorded for origin, course, and exit site. Most were imaged with parasternal long and short axis views from the apical window. The subxiphoid window also was used in younger patients. The coronary artery diameter was considered normal if measurements were within two SDs of normal controls. Left ventricular dimensions were assessed two dimensionally, and functions were assessed according to the recommendations for quantification methods during the performance of a pediatric echocardiogram.[8]

Cost-effectiveness

The charge levied for each examination of asymptomatic fistula (with or without echocardiography) was derived from the state record or private health insurance company invoice files.

RESULTS

Clinical findings at presentation

Among the 78,000 patients who had had an echocardiographic examination at outpatient clinics during the 12-year period covered, 118 (0.053%) had been found by color Doppler flow mapping to have a clinically silent CAF. A total of 300 echocardiograms had been performed on these patients. The demographic, clinical, follow-up, and echocardiographic data are presented in Table 1.

Table 1

Clinical, echocardiographic findings, and outcome data in 118 patients with clinically silent coronary artery fistulae

Patient no	118	118
Age at diagnosis (months)	Mean±SD (Min-max)
	70.8±73.7 (0-23)
	Median: 50 IQR 13.2-135
Sex	Female: 56 (47.5%)	Male: 62 (52.5%)
Weight (kg)	Mean±SD (min-max)
	23.5±20.9 (2-69)
	Median: 9.5 IQR 6-38
Follow-up time (months)	Mean±SD (min-max)
	41.3±34.6 (1-120)
	Median: 41.3 IQR 26-50
Reason for ECHO	Murmur, chest pain, control
Fistula diameter (mm)	Mean±SD (Min-max)
	1.4±0.4 (1-3)
	Median: 1.4 IQR 1.2-1.6
Exit Site	MPA1: 26
	MPA2: 76
	LPA: 2
	RPA: 12
	RVOT: 2
Echocardiographic findings	LVDD (mm) Mean±SD (min-max) 33.6±11.7 (7-51)	EF (%)
		Mean±SD (min-max)
		69.4±4.7 (52-82)

MPA1: fistulous flow located in anterior aspect of main pulmonary artery (between pulmonary valve and bifurcation), MPA2: fistulous flow located in aortic side of main pulmonary artery, RPA: right pulmonary artery, LPA: left pulmonary artery, RVOT: right ventricular outflow tract, LVDD: Left ventricular end diastolic diameter, EF: ejection fraction. ECHO: Echocardiography

The mean age at presentation was 70.8 ± 73.7 months (range: 0–223 months). The initial indication for pediatric echocardiography at presentation was murmur (77 patients), suspicion of congenital heart disease (19 patients), and systolic murmur during routine physical examination (23 patients). Associated congenital cardiac abnormalities were present in 33 of 118 patients [Table 2].

Table 2

Co-existent cardiac findings with fistulas

Co-existent cardiac findings	No. of Patients	%
None	35	29.7
AI	6	5.1
AS, VSD	1	0.8
ASD	2	1.7
ASD, MI	2	1.7
MI	9	7.6
MI, AI	2	1.7
MVP	2	1.7
MVP, MI	3	2.5
MVP, MI, AI	2	1.7
PDA	1	0.8
PFO	48	40.7
VSD	3	2.5
VSD PDA	1	0.8
VSD, AI	1	0.8
Total (n)	118	100.0

AI: Aortic insufficiency, AS: Aortic valve stenosis, ASD: Atrial septal defect, MI: mitral insufficiency, MVP: Mitral valve prolapses, PDA: Patent ductus arteriosus, PFO: patent foramen ovale, VSD: Ventricular septal defect.

Echocardiographic findings

The left ventricular end-diastolic dimensions were in the normal ranges for all patients at presentation, while ejection fractions were in the normal ranges for all patients at both presentation and follow-up. The mean ejection fraction was 69.4 ± 4.7 (52–82). The origin of the fistulae was not clearly defined by color Doppler echocardiography except in two of the patients (in whom the fistulous flow had come from the right coronary artery to the right ventricle). However, the exit point was clearly visualized with color Doppler in all the patients. No patients had diagnostic catheter angiocardiography. The fistulae drained to different sites in the pulmonary arteries. In most patients, the CAF was located around the pulmonary valve and the bifurcation; in 76 patients, it was on the aortic side of the main pulmonary artery, 26 patients had fistulous flow through the anterior wall of the main pulmonary artery, two had the right ventricular outflow tract (RVOT) as the exit site, for 12, it was the right pulmonary artery, and for another two, the left pulmonary artery [Figure 1]; in 12, it was in the right pulmonary artery; and in another two patients, the fistulae drained into the left pulmonary artery [Figure 2]. Additional cardiac comorbidities are presented in Table 2.

Figure 1

(a-d) Fistulous flow located in main pulmonary artery between pulmonary valve and bifurcation. Arrow shows the opening of the fistula; RPA: Right pulmonary artery, LPA: Left pulmonary artery

Figure 2

(a, b and d) Fistulous flow tract located in main pulmonary artery beneath the pulmonic valve; (c) Fistulous flow tract drained to the right ventricular outflow in main pulmonary artery. Arrow shows the opening of the fistula; RPA: Right pulmonary artery, LPA: Left pulmonary artery

Follow-up

The 118 patients had ongoing follow-up (mean age: 41 months range: 1 month to 10 years). There was no spontaneous fistula resolution. The patients were asymptomatic without intervention and with ongoing echocardiographic evidence of small CAF at the last follow-up. They had no complaint or sign of chest pain, exercise intolerance, or electrocardiographic abnormality.

Cost-effectivity

Patients had paid 50 US dollars for each echocardiographic examination at state hospitals and 150 through private health insurance. A total of 354 echocardiograms were performed, 240 at state hospitals and 114 at private practices.

DISCUSSION

The widespread availability of echocardiographic imaging with color Doppler flow mapping in pediatric cardiology reveals clinically silent findings.[9] The incidence of CAF reported ranges from 0.2% to 0.4% in congenital heart disease patients and approximately 0.2% in adults undergoing coronary angiography.[1011]

Among the 32,385 patients who had an echocardiographic examination at a children's hospital in Boston during a 10-year period, a 0.06% incidence of clinically silent CAF was found by color Doppler.[9] This was reported as an underrepresentation of the true incidence because of missing complete Doppler results in some subjects.

In this review study, we found 118 young patients with clinically silent CAF revealed through color Doppler flow mapping at an incidence of 0.015% of all patients that had echocardiography with irrelevant causes, which is less than the Boston results. We cannot estimate the incidence of CAF in the overall pediatric population since the study population was not part of a screening program.

Although all the patients were referred to the pediatric cardiology consultant because of a murmur, only 13 had a murmur relevant to their concomitant cardiac defect. Of the 118 patients found by color Doppler to have a clinically silent CAF, 13 had a holosystolic murmur, which is best heard at the apex/left sternal border (due to the mitral insufficiency/ventricular septal defect), and four patients had a soft mid-systolic murmur at the upper left sternal border (with wide and fixed splitting of the second heart sound) (S2). It is reported that associated congenital cardiac anomalies are evident in 20%–45% of pediatric cases of CAF.[12]

Various cardiac anomalies have been reported in the published coronary fistula series. Holzer et al. found five of 17 (29%) patients to have additional cardiac anomalies along with CAF. They summarized the (English) literature as showing additional cardiac anomalies in 106 (25%) of 426 pediatric patients with coronary fistulas. Atrial septal defects and ventricular septal defects were among the most common associated anomalies.[13] A review by Said et al. found an incidence of additional cardiac anomalies in 21% (range: 5%–30%) of pediatric populations with CAF.[14] In this study, we had 35 (33%) patients with associated heart defects. The incidence of associated heart defects here was higher than in Said's study mostly because of our extended heart defect spectrum. We included all valve insufficiencies, including mild ones.

Most symptomatic CAF originates at the right coronary artery system and drains to the right ventricle, right atrium, or main pulmonary artery.[15] In this study, we found two patients whose CAF originated at the right coronary artery system and drain into the right ventricle; these patients were asymptomatic. In most of the patients, the fistulous flow was found to be located around the pulmonary valve and the pulmonary bifurcation; specifically, in 76 of these patients, fistulous flow was on the aortic side of the main pulmonary artery, in 26, it was through the anterior wall of the main pulmonary artery, and in two patients, the exit site was at the RVOT. These proportions are similar to those reported. In two patients, the fistula drained into the left pulmonary artery, which is very rare.[16]

The incidental diagnosis of small fistulae upon color Doppler echocardiography has raised concerns about the management of these fistulae. The American College of Cardiology/American Heart Association guideline recommends percutaneous or surgical closure for large fistulae where there is evidence of myocardial ischemia, arrhythmia, or ventricular dysfunction enlargement in small-to-moderate-sized fistulae. Incidentally detected, small asymptomatic fistulae do not require closure in adults.[17] There is no recommendation for children with incidentally detected asymptomatic fistulae.

In this study, we identified 118 patients with fistulae diagnosed incidentally during echocardiography, the patients having been directed to the pediatric cardiology consultant for other (irrelevant) reasons. Mostly, these children were from high socioeconomic level families with private health insurance. Assessment of the importance of the diagnosis and information on outcomes is the responsibility of the consultant physician. Although small, asymptomatic fistulae do not require a specific procedure for closure when detected incidentally without any symptom,[18] due to the proximity of the pulmonary artery–coronary arteries, they are mostly thin coronary pulmonary artery fistulae.[1920]

Said et al. have reported conservative medical management for asymptomatic small CAFs. They stated that asymptomatic CAFs are benign in long-term follow-up and may demonstrate spontaneous closure, finding that complications had not occurred in either the pediatric (36%) or adult (24%) populations with conservative medical management. Sherwood et al. concluded that incidentally detected, small CAFs in asymptomatic children and young adults found by color Doppler echocardiography could be followed noninvasively and without intervention. Their findings are in agreement with those of Hobbs et al., who found no new adverse developments in 101 adults with small CAF identified by angiography (after following the patients clinically for up to 11 years).[21]

Reports are in general agreement also with regard to recommending closure of the fistula in certain situations. These include those cases in which the patients are considered to be at high risk for later infective endocarditis, where they are not feasible for longitudinal follow-up and where they will have an invasive procedure for some other cardiac problem.[1822]

Complications in congenital CAFs, such as aneurysmal formation, rupture of the fistula, and thrombus formation, are mostly reported among adults; the low incidence in pediatric patients may reflect the influence of age on the fistulae. During the mean 41-month-long follow-up following a mean age of presentation of 70 months, we had neither clinical symptoms associated with our CAF patients nor echocardiographic findings indicating fistula dilation, aneurysm, or thrombosis. The mean follow-up period was not stated in the review in Said et al.'s study, which reported aneurysmal formation at 9%.[14]

Follow-up of asymptomatic small CAF patients has both echocardiography and downstream costs for additional checkups. On the other hand, early asymptomatic CAF detection may require a specific procedure for closure when detected incidentally without any symptom. Currently, there is no such asymptomatic incidentally diagnosed CAF follow-up program for children in the world. We therefore assessed the long-term health effect and costs of follow-up programs in CAF detected in children.

Dealing with a CAF is not unlike the parable of the blind men describing an elephant. Reports in the literature are mostly based on the authors' limited experience, and their descriptions of CAFs differ widely because of the rarity and unpredictability of the condition in the growing child. In some papers, antibiotic prophylaxis is strongly recommended for both pediatric and adult patients with congenital CAFs,[23] while in others, a conservative follow-up is strongly suggested for patients with nonsignificant CAF, for whom it is reported, intervention procedures may be unnecessary.[24]

CONCLUSION

Given that neither a lack of clinical symptoms nor echocardiographic stability of CAF over a 5-year follow-up period delineates a lifelong nonoccurrence of fistula, we can only recommend that these patients be followed regularly for their CAF. In our practice, we have used transthoracic echocardiogram in the investigation and follow-up of pediatric CAF. We recommend that the treatment and follow-up strategy for pediatric patients be individualized according to fistula origin, size, chamber enlargement, drainage site, patient age, and cost-effectiveness.

Ethical clearance

Informed consent was obtained from the parents or legal guardians of all individual participants included in the study.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.