Outcomes of Transcatheter Closure of Congenital Left Circumflex Coronary Artery Fistula

Peijian Wei; Fang Fang; Fengwen Zhang; Yihang Li; Pengxu Kong; Shuyi Feng; Zhongying Xu; Liang Xu; Junyi Wan; Gejun Zhang; Xiangbin Pan

doi:10.1253/circj.CJ-23-0800

Abstract

Background: Congenital left circumflex coronary artery fistula (LCX-CAF) is a relatively rare type of coronary artery fistula (CAF); little is known about the outcomes of transcatheter closure (TCC) of LCX-CAF.

Methods and Results: All consecutive patients admitted to Fuwai Hospital and scheduled for TCC of LCX-CAF between January 2012 and December 2022 were reviewed retrospectively. Of the 25 consecutive patients (mean [±SD] age 34±20 years; 48% male) admitted and scheduled for TCC of congenital LCX-CAF, the procedure was feasible in 22 (77.3%). The mean (±SD) diameter of the fistulas was 6.99±2.04 mm; 21 (84%) patients had a large fistula (i.e., diameter >2-fold greater than non-feeding coronary artery). Occluders were deployed via a transarterial approach and arteriovenous loop in 6 (27.3%) and 16 (72.7%) patients, respectively. No procedural complications were recorded. Although the procedural success rates are similar for single LCX-CAF and left anterior descending CAF (81.25% vs. 92.86%; P=0.602), the mean time from initial angiography to first occluder deployment is significantly longer for LCX-CAF (83.06±36.07 vs. 36.00±9.49 min; P<0.001). The mean (±SD) follow-up time was 62.2±45.5 months. The incidence of myocardial infarction and recanalization of the fistula was 4.5% (1/22) and 9.1% (2/22), respectively.

Conclusions: TCC of LCX-CAF is a feasible and effective alternative to surgical repair, with comparable outcomes in selected patients. Optimal medical therapy to prevent post-closure myocardial infarction requires further investigation.

Congenital coronary artery fistulas (CAFs) are abnormal connections between the coronary arteries and major blood vessels or cardiac chambers. It has been reported that CAFs were present in 0.002–0.3% of patients who underwent invasive coronary angiography and in 0.9% of patients who underwent computed tomographic angiography.¹ Left circumflex coronary artery fistula (LCX-CAF) is a relatively rare type of CAF, with an incidence of 5–20%.² Transcatheter closure (TCC) of CAFs is associated with low morbidity and a favorable procedural success rate in selected patients with suitable anatomy,³ and has recently emerged as a relatively common strategy. However, TCC of LCX-CAF is technically challenging due to the characteristics of the circumflex artery course. Despite there being a few reported cases,⁴^,⁵ information is limited regarding the outcomes of TCC of LCX-CAF. Thus, the aim of the present study was to describe outcomes following TCC of LCX-CAF and review our experience.

Methods

This retrospective study was approved by the Ethics Committee of Fuwai Hospital, Chinese Academy of Medical Sciences (No. 2023-1126). The requirement for individual informed consent was waived due to the retrospective nature of the study and the use of anonymized data.

Patients

A retrospective review was performed of all consecutive patients who were admitted to the Department of Structural Heart Disease of Fuwai Hospital and scheduled for TCC of LCX-CAF between January 2012 and December 2022. Patient demographics, clinical manifestations, perioperative and follow-up electrocardiogram, chest radiograph, perioperative and follow-up transthoracic echocardiograms, computed tomography coronary angiography (CTCA), and selective coronary angiography were obtained from the medical records database. All data were anonymized before processing. The anatomic characteristics of LCX-CAF were further confirmed by reviewing both CTCA and selective coronary angiography.

LCX-CAF was defined as large, medium-sized, or small if the diameter of the fistula was >2-fold greater, 1- to 2-fold greater, or less than that of the non-feeding coronary artery, respectively.⁶ Coronary artery aneurysms were defined as dilated coronary arteries with a diameter >1.5-fold greater than the diameter of normal coronary arteries, whereas giant coronary artery aneurysms were defined as those with a diameter >20 mm.⁷ Given that CAFs most commonly originate from the left anterior descending coronary artery (LAD),⁸ we further compared the procedural outcomes of patients with single LAD-CAF and LCX-CAF over the same period to determine whether the procedural success rate of TCC for LCX-CAF is indeed lower.

Procedural Technique

Local anesthesia was used, except in the case of 3 pediatric patients. Vascular access was established via the right femoral artery and heparin was then administered (100 units/kg). Initial selective coronary angiography was performed to further determine the anatomy of LCX-CAF and to facilitate selection of occluder type, size, and position for deployment.

For patients with medium or large fistulas draining into low-pressure chambers or veins and the pulmonary artery (Figure 1A–O), vascular plug, ductal occluders, or ventricular septal occluders were preferred and deployed via an arteriovenous loop (Figure 2). It was noted that the key was to advance the guidewire through the drainage site. If this failed, the coils were then deployed at the narrow and distal portion of the fistula.

Figure 1.

Types of left circumflex coronary artery fistulas: (A–C) left circumflex coronary artery (LCX) to pulmonary artery (PA) fistula; (D–F) LCX to superior vena cava (SVC) fistula; (G–I) LCX to right atrium (RA) fistula; (J–L) LCX to right ventricle (RV) fistula; (M–O) LCX to coronary sinus (CS) fistula; (P–R) LCX to left ventricle (LV) fistula. LAD, left anterior descending coronary artery; RCA, right coronary artery.

Figure 2.

Transcatheter closure of a left circumflex coronary artery (LCX) fistula draining into the right atrium (RA) using the wire-maintaining technique. (A) Selective coronary angiography was performed to further determine the anatomy of the fistula. (B,C) The 0.01400 coronary guidewire was advanced to the LCX, RA, right ventricle, and inferior vena cava, and was then snared and exteriorized from the right femoral vein to establish an arteriovenous loop. (D) A patent ductus arteriosus occluder (Lifetech Scientific, Shenzhen, China) was advanced through the delivery sheath while the guide wire arteriovenous loop was maintained in situ. (E) Selective coronary angiography was performed before the detachment of the occluder to further confirm the appropriate positioning of the device by assessing the presence of residual flow. (F) The occluder was deployed.

For patients with fistulas draining into the left ventricle (Figure 1P–R), TCC of the fistulas was performed using previously described techniques.⁹ Repeated angiography of the fistula was performed immediately after occluder deployment to determine whether there was residual shunt. Residual shunt was classified as trivial (shunt <1 mm), mild (shunt 1–2 mm), moderate (shunt 3–4 mm), or severe (shunt >4 mm). Hemodynamics and 12-lead electrocardiography (ECG) were monitored for 10–15 min as an occlusion test.

After the procedure, anticoagulation therapy or antiplatelet agents were prescribed for all patients for at least 6 months.

Outcomes and Follow-up

The primary outcomes in this study were procedural success and procedural complications: myocardial infarction, ischemic changes on ECG, coronary artery dissection, coronary artery spasm, and device embolization. Secondary outcomes included improvement of symptoms and long-term complications, including recanalization of the fistula and reintervention. Procedural success was defined as no or trivial residual flow on selective coronary angiography without procedural complications. Follow-up information was obtained from visits to the outpatient clinic and/or by telephone investigation at 1, 6, and 12 months after the procedure and annually thereafter.

Statistical Analysis

Categorical variables are reported as frequency and percentage, and continuous variables are presented as the mean±SD or median (range). Categorical variables were compared using the Chi-squared test or Fisher’s exact test as appropriate, and continuous variables were compared using an independent samples t-test. Two-sided P<0.05 was considered statistically significant. Statistical analyses were conducted using IBM SPSS Statistics 25 (IBM Corp., Armonk, NY, USA).

Results

Baseline Characteristics

Between January 2012 and December 2022, 25 consecutive patients were admitted to Fuwai Hospital and scheduled for TCC of congenital LCX-CAF. The detailed demographic characteristics of these patients are presented in Table 1. Of the 25 patients, 48% were male and 52% were female, the mean age was 34±20 years, and the mean body mass index was 21.84±3.61 kg/m². Regarding clinical manifestations, 5 patients presented with palpitation, 5 with dyspnea on exertion, 10 with chest pain, and 8 with murmur. Increased pulmonary perfusion was detected in 17 (68%) patients on chest radiographs.

Table 1.

Baseline Characteristics of 25 Patients With Left Circumflex Coronary Artery Fistulas

Demographic features
Age (years)	34±20
Age (<18 years)	3 (12)
Sex
Male	12 (48)
Female	13 (52)
BMI (kg/m²)	21.84±3.61
Clinical manifestation
Asymptomatic	9 (36)
Palpitation	5 (20)
Dyspnea on exertion	5 (20)
Chest pain	10 (40)
Murmur	8 (32)
Comorbidities
Hypertension	2 (8)
Coronary artery disease	1 (4)
Atrial septal defect	1 (4)
Patent foramen ovale	1 (4)
Electrocardiogram
Normal	10 (40)
ST segment or T wave changes	7 (28)
LV high voltage	7 (28)
Atrial premature beats	1 (4)
Chest radiograph
Increased pulmonary perfusion	17 (68)
Cardiothoracic ratio	0.50±0.06
Echocardiogram
Fistula diameter (mm)	6.99±2.04
LV ejection fraction (%)	0.63±0.07

Data are given as the mean±SD or n (%). BMI, body mass index; LV, left ventricular; SD, standard deviation.

The anatomic characteristics of the fistulas are presented in Table 2. Six (24%) patients had multiple origins of the fistula. In terms of the location of origin relative to the fistula, it was proximal in 10 (44%) patients and distal in 14 (56%) patients. The draining sites included the superior vena cava (n=1; 4%), coronary sinus (n=5; 20%), right atrium (n=9; 36%), right ventricle (n=4; 16%), left ventricle (n=4; 16%), and pulmonary artery (n=2; 8%; Figure 1). The most common pathway morphology of the fistula was a single origin, single draining site with a tortuous pattern (n=18; 72%). Coronary aneurysms and giant coronary aneurysms were present in 15 (60%) and 10 (40%) patients, respectively. The mean diameter of the fistulas was 6.99±2.04 mm, and most fistulas (n=21; 84%) were large.

Table 2.

Anatomic Characteristics of Left Circumflex Coronary Artery Fistulas in 25 Patients

Origin site
LCX	19 (76)
RCA+LCX	1 (4)
LAD+LCX	4 (16)
LAD+RCA+LCX	1 (4)
Multiple origins	6 (24)
Proximity of origin
Proximal	10 (44)
Distal	14 (56)
Draining site
Superior vena cava	1 (4)
CS	5 (20)
Right atrium	9 (36)
RV	4 (16)
Left ventricle	4 (16)
PA	2 (8)
Pathway morphology of the fistula
Single origin, single draining site, and straight	2 (8)
Single origin, single draining site, and tortuous	18 (72)
Single origin, multiple draining sites, and tortuous	1 (4)
Multiple origins, single draining site, and tortuous	2 (8)
Multiple origins, multiple draining sites, and tortuous	2 (8)
Aneurysm formation	15 (60)
Giant aneurysm formation	10 (40)
Fistula size
Fistula diameter (mm)	6.99±2.04
Medium^A	4 (16)
Large^B	21 (84)

Data are given as the mean±SD or n (%). ^AA medium-sized fistula was defined as one with a diameter that was 1- to 2-fold greater than that of the non-feeding coronary artery. ^BA large fistula was defined as one with a diameter that was >2-fold greater than that of the non-feeding coronary artery. CS, coronary sinus; LAD, left anterior descending branch; LCX, left circumflex branch; PA, pulmonary artery; RCA, right coronary artery; RV, right ventricle; SD, standard deviation.

Procedural Outcomes

In 3 patients, several attempts to advance the guidewire into the fistula were unsuccessful due to the extreme tortuosity and large angle torsion of the fistula. Given that the fistula in these patients was >5 mm in diameter with a dilated course, coil embolization of the fistula via the arterial approach was unsuitable,¹⁰ and eventually the procedure was stopped. Thus, the procedure was feasible in only 22 of 25 (88%) patients scheduled for TCC of congenital LCX-CAF.

Detailed procedural outcomes for the 22 patients who ultimately underwent TCC are presented in Table 3. The mean time from initial angiography to first occluder deployment was 79.95±35.94 min. Occlusion devices included a ventricular septal defect occluder (n=1; 4.5%), a vascular plug (n=3; 13.6%), coils (n=4; 18.2%), and a patent ductus arteriosus occluder (n=14; 63.6%). The occluders were deployed via the anterograde arterial approach, retrograde arterial approach, and arteriovenous loop in 5 (22.7%), 1 (4.5%), and 16 (72.7%) patients, respectively. No/trivial residual shunt was observed in 18 (81.8%) patients, and mild residual shunt was observed in 4 (18.2%) patients.

Table 3.

Procedural Outcomes for 22 Patients With Left Circumflex Coronary Artery Fistulas Who Ultimately Underwent Transcatheter Closure

Patient no.	Course	Pathway morphology	Approach	Device	Size	Residual shunt
1	LCX→RA	Single draining site and tortuous	Arteriovenous loop	Vascular plug^A	6 mm/6 mm	Mild
2	LCX→CS	Single draining site and tortuous	Arteriovenous loop	Vascular plug^A	6 mm/6 mm	No
3	LCX→RA	Single draining site and tortuous	Arteriovenous loop	PDA occluder^B	10 mm×12 mm	No
4	LAD+LCX→RA	Single draining site and tortuous	Arteriovenous loop	PDA occluder^B	10 mm×12 mm	Trivial
5	LCX→RV	Single draining site and straight	Arteriovenous loop	PDA occluder^B	8 mm	No
6	LCX→RV	Single draining site and tortuous	Arteriovenous loop	PDA occluder^B	10 mm×12 mm	No
7	LCX→RA	Single draining site and tortuous	Arteriovenous loop	PDA occluder^B	10 mm×12 mm	Trivial
8	LCX→LV	Single draining site and tortuous	Retrograde arterial	PDA occluder^B	16 mm×18 mm	No
9	LCX→RV	Single draining site and straight	Arteriovenous loop	PDA occluder^B	8 mm×10 mm	Mild
10	LAD+LCX→RV	Single draining site and tortuous	Arteriovenous loop	Vascular plug^A	4 mm/6 mm	No
11	LCX→RA	Single draining site and tortuous	Arteriovenous loop	PDA occluder^B	8 mm×10 mm	No
12	LCX→RA	Single draining site and tortuous	Arteriovenous loop	PDA occluder^B	6 mm×8 mm	No
13	RCA+LCX→LV	Single draining site and tortuous	Anterograde arterial	Cook coils^C	8 mm×20 cm, 6 mm×10 cm	No
14	LCX→CS	Single draining site and tortuous	Arteriovenous loop	PDA occluder^B	18 mm×20 mm	Trivial
15	LCX→RA	Multiple draining sites and tortuous	Arteriovenous loop	PDA occluder^B	8 mm×10 mm*2	No
16	LCX→CS	Single draining site and tortuous	Arteriovenous loop	PDA occluder^B	20 mm×22 mm	Mild
17	LCX→SVC	Single draining site and tortuous	Arteriovenous loop	PDA occluder	10 mm×12 mm	No
18	LCX→LV	Single draining site and tortuous	Anterograde arterial	PDA occluder	20 mm×22 mm	Mild
19	LAD+RCA+LCX→PA	Multiple draining sites and tortuous	Anterograde arterial	Cook coils^C	8 mm×20 cm*2	No
20	LAD+LCX→LV	Single draining site and tortuous	Anterograde arterial	Cook coils^C	6 mm×10 cm	No
21	LAD+LCX→PA	Multiple draining sites and tortuous	Anterograde arterial	Cook coils^C	3 mm×6 cm*2, 2 mm×3 mm×2.3 cm	No
22	LCX→CS	Single draining site and tortuous	Arteriovenous loop	VSD occluder^D	12 mm	No

^AAmplatzer duct occluder II (AGA Medical Corporation, Plymouth, MN, USA). ^BPDA occluder (Starway Medical Technology [Beijing, China], Lifetech Scientific [Shenzhen, China], and Lepu Medical Technology [Beijing, China]). ^CInterlock coils (Boston Scientific Corporation, Marlborough, MA, USA). ^DVSD occluder (Lifetech Scientific). PDA, patent ductus arteriosus; RA, right atrium; SVC, superior vena cava; VSD, ventricular septal defect. Other abbreviations as in Tables 1,2.

Given that the deployed device affected coronary venous blood return and caused pericardial effusion, occluder retrieval was performed in a patient (Patient no. 2) with a fistula draining into the coronary sinus on the second postoperative day even though there was no residual shunt after deployment. Thus, the procedural success rate was 77.3% (17/22). The median postoperative in-hospital length of stay was 4 (1–10) days. The incidence of transient ST-T segment change on ECG was 18.1% (n=4). With regard to procedural complications, no deaths, coronary artery dissections, myocardial infarctions, or device embolization were recorded.

Follow-up

The mean follow-up time was 62.2±45.5 months (range 7.5–149.8 months). Two patients experienced recanalization of the fistula. Of these patients, 1 who was asymptomatic was followed up further, whereas the other (Patient no. 1) with mild residual shunt at discharge successfully underwent repeated TCC using a 6/6-mm Amplatzer Duct Occluder II (AGA Medical Corp., Plymouth, MN, USA) 1 month after the procedure due to the rapid progression of residual shunt and unimproved symptoms.

One patient who received postoperative warfarin alone was readmitted to the emergency department 1 month after the procedure due to acute chest pain, with an ECG revealing inferior posterior myocardial infarction. CTCA images further confirmed the presence of coronary artery thrombosis extending from the fistulous tract. A significant improvement in symptoms was observed after the patient was treated with a combination of aspirin, clopidogrel, and warfarin. Of note, complete closure of the fistula and thrombosis of the fistulous tract adjacent to the occluder were observed on CTCA images 1 month after the procedure in another 2 patients. In addition, CTCA images 1 year after the procedure obtained from 6 patients demonstrated total closure of the fistula in all patients, with thrombotic occlusion of the fistulous tract with a patent feeding coronary artery in 3 patients and patency of the fistulous tract without thrombosis in the other 3 patients.

According to the last clinical evaluation and/or telephone investigation, most (15/22; 68.9%) patients were in New York Heart Association Class I. Specifically, bradycardia was detected on ECG in 2 patients, and chest tightness was reported by 3 patients.

Comparison of Procedural Outcomes Between LCX-CAF and LAD-CAF

During the same period, TCC was attempted for single LAD-CAF in 15 patients, with the procedure feasible in 14 (93.33%). Baseline data and anatomic characteristics of the fistula for patients with single LAD-CAF (n=14) and single LCX-CAF (n=16) who ultimately underwent TCC are presented in Supplementary Tables 1,2. The procedural success rate for LAD-CAF was higher than that for LCX-CAF (92.86% vs. 81.25%; P=0.602), although the difference was not statistically significant. However, the mean time from initial angiography to the first deployment of the occluder was significantly shorter for LAD-CAF patients (36.00±9.49 vs. 83.06±36.07 min; P<0.001). The mean follow-up duration for LAD-CAF patients was 25.72±23.63 months, with similar recanalization (LAD-CAF: 7.14%; LCX-CAF: 12.5%; P=1) and reintervention (LAD-CAF: 0%; LCX-CAF: 6.25%; P=1) rates to those of LCX-CAF patients.

Discussion

Challenges Posed by Anatomical Features of the LCX

According to the 2020 ESC guidelines, the indications for surgical closure or TCC in patients with CAFs include the presence of symptoms, complications, and significant shunting.¹¹ Considering that TCC can avoid complications associated with median thoracotomy and cardiopulmonary bypass and has been proven to be feasible with comparable outcomes, it is preferred over surgical closure in patients with suitable anatomy.¹² As a relatively rare type of CAF, LCX-CAF poses a challenge for both surgical repair and TCC due to the tortuous vessel course. In general, the circumflex artery arises from the bifurcation of the left main coronary artery and supplies blood to the posterolateral left ventricle and left atrium. It runs along the left atrioventricular groove, first to the left rear around the heart, and then to the right, generally reaching near the posterior wall of the left ventricle. However, it is widely acknowledged that cardiopulmonary bypass should be available when the fistula is in the left atrioventricular groove or on the posterior surface of the heart.¹³ Thus, surgical repair of LCX-CAF is more likely to be performed under cardiopulmonary bypass. According to a previous study by Hou et al, only 4 patients underwent off-pump suture ligation among 29 LCX-CAF patients who underwent surgical repair.¹⁴

Limited Data for TCC of LCX-CAF

In terms of TCC of LCX-CAF, the procedure is technically challenging due to vessel tortuosity and sharp angulations of the LCX that make navigating the guidewire into the fistula difficult. Fortunately, on the strength of improvements in the deliverability of devices and larger experience, there are reports of the use of TCC in some LCX-CAF patients with technically challenging anatomy.⁴^,⁵ Thus far, however, there has been no single study that has systematically reviewed the feasibility, skills required, and procedural outcomes of TCC of LCX-CAF. To date, this is the only study to report on the procedural outcomes and experience of a series of TCC of LCX-CAF.

In this study, 3 of the 25 (12%) patients who were deemed suitable for TCC after a preoperative assessment based on angiographic or CTCA images eventually failed to undergo TCC after several failed attempts; thus, TCC of LCX-CAF was not feasible for these 3 patients. The procedural success rate of TCC of LCX-CAF among the other 22 patients was 77.3% (17/22), which is relatively low compared with the success rate of TCC of congenital CAFs (87.5%; 42/48) reported in a large single-center study.³ Upon further comparison of procedural outcomes among patients with single LAD-CAF and LCX-CAF who underwent TCC, it was observed that the success rate of TCC for LAD-CAF was higher than that for LCX-CAF (92.86% vs. 81.25%; P=0.602). Moreover, the mean time from initial angiography to first deployment of the occluder was significantly shorter for LAD-CAF (36.00±9.49 vs. 83.06±36.07 min; P<0.001). This further confirms the difficulty of manipulating and advancing the guidewire to the planned deployment position in a tortuous fistulous tract.

Given the excellent deliverability of coils, they are commonly used in TCC of CAFs with tortuous courses. In previous studies, embolization coils accounted for 70.8–90% of devices used for TCC of CAFs.³^,¹⁵^,¹⁶ Recently, Abdelfattah et al reported successful angiographic transcatheter coil embolization in 88% of CAF cases.¹⁷ In the present study, tortuous fistulas were observed in 91% (20/22) of LCX-CAF patients. However, coils were used for TCC in only 18.2% (4/22) of cases. Specifically, transcatheter coil embolization should take into account the size of the fistula and the dilation of the coronary artery due to the risk of coil migration or embolization. Transcatheter coil embolization is preferred in small or medium-sized fistulas.¹⁰ In our series, the mean diameter of the fistulas was 6.99±2.04 mm, and 84% (n=21) of fistulas were classified as large. Therefore, transcatheter coil embolization was less common in the TCC of LCX-CAF, whereas ductal occluders were the most used.

Experience With TCC of LCX-CAF

Regarding procedural skills, TCC of CAFs can be performed via a transarterial approach, a transvenous approach, or arteriovenous loop.⁶ Interestingly, no device was deployed via a transvenous approach in the present study. The reason for this is that an arteriovenous loop is preferably used to provide extra support for navigating and advancing the occluder into the extremely tortuous fistula. Therefore, occluders were placed via the transarterial approach and arteriovenous loop in 6 (27.3%) and 16 patients (72.7%), respectively. Moreover, the longer time from initial angiography to first occluder deployment may also be associated with the higher proportion of patients in whom the arteriovenous loop was used. Zhang et al proposed that wire-maintaining techniques can facilitate TCC of medium and large complex CAFs.¹⁸ In the present series, the wire-maintaining technique was used in 3 patients to avoid potential vessel injury with repeated device placement and thus avoid further prolongation of the procedure. In addition, a transcatheter stent-assisted coil occlusion technique,¹⁹ a “child in mother” technique with dedicated catheters via the transradial approach,²⁰ and a novel telescope anchoring technique using a microcatheter, guide extension, and balloon anchoring²¹ have been described to facilitate successful TCC of LCX-CAF and can serve as a reference.

In determining the optimal occlusion site, the fundamental principle is to attain complete closure of the fistula while safeguarding normal coronary flow. For fistulas with a proximal-type origin, an occlusion site may be either proximal to the origin or distal to the drainage site, aiming to minimize the risk of thrombosis.²² For fistulas with a distal-type origin, it is suggested that the fistula is closed as close as possible to the draining site while avoiding any inadvertent occlusion of branches supplying the normal myocardium.¹⁰

Concerns About Procedural Complications

The combination of CAFs and coronary artery aneurysms was considered rare, with coronary artery aneurysms observed in 15% of CAF patients and giant coronary artery aneurysms seen in only 5.9% of CAF patients.²³ Recently, Ouchi et al found that the incidence of aneurysm accompanying CAF determined by CTCA was 48.4%.²⁴ In the present study, coronary aneurysms and giant coronary aneurysms were present in 15 (60%) and 10 (40%) patients, respectively. This suggests that aneurysm formation is more common in patients with LCX-CAF. Due to aneurysm complications of thrombosis, embolization, and rupture, the management of CAFs patients with coronary aneurysm is controversial and includes TCC and surgical repair.²⁵ In addition, the presence of a coronary aneurysm is associated with a high risk of thrombus formation secondary to stagnant flow after both TCC and surgical closure of CAFs.⁶ Further, a proximal fistula closed distally is linked to an increased likelihood of thrombosis because the sudden cessation or significant reduction in blood flow velocity within the post-closure fistulous tract may lead to thrombosis.²⁶

A computational fluid dynamics study by Cao et al suggested that the tortuous geometry of CAFs directly contributes to worse hemodynamics within the fistula, which deteriorates following terminal closure.²⁷ Of note, in our series the pathway morphology of the fistula was tortuous in 72% of patients with LCX-CAF. Therefore, concerns regarding the risk of myocardial infarction after TCC of LCX-CAF were more widespread, especially in patients with drainage into the coronary sinus. Valente et al reported that CAFs that drain into the coronary sinus were the sole angiographic feature predictive of long-term ischemic events.²⁸ In the present study, there were 4 patients with an LCX-CAF draining into the coronary sinus, a rare type of CAFs with limited reports.²⁹ One of these patients experienced myocardial infarction 1 month after closure while on therapeutic warfarin alone. Interestingly, clopidogrel and aspirin were prescribed in all patients except this patient for at least 6 months. Importantly, optimal medical therapy to prevent myocardial infarction after closure remains unknown. Moreover, in another 2 patients, complete closure of the fistula and thrombosis of the fistulous tract adjacent to the occluder were observed on CTCA images 1 month after the procedure. These findings suggest that thrombosis may be more common after closure of LCX-CAF and antiplatelet therapy may prevent post-closure myocardial infarction. In addition, the incidence of recanalization of the fistula was 9.1%. Comparable outcomes suggest that TCC of LCX-CAF is a feasible and effective alternative to surgical repair.

Study Limitations

First, this was a single-center retrospective study with a limited sample size owing to the rarity of LCX-CAF. However, to the best of our knowledge, it is the only study focusing on the procedural outcomes of TCC of LCX-CAF. Second, follow-up CTCA was completed in a few patients. The incomplete CTCA data led to uncertainty regarding the incidence of thrombotic occlusion of the fistulous tract.

Conclusions

For selected patients, TCC of LCX-CAF is a feasible and effective alternative to surgical repair with comparable outcomes. However, thrombosis may be more common after closure of LCX-CAF, and optimal medical therapy to prevent post-closure myocardial infarction requires further investigation.

Acknowledgment

None.

Sources of Funding

This work was supported by grants from CAMS Innovation Fund for Medical Sciences(2021-I2M-1-065), National Key R&D Program of China (2022YFC2503400, 2023YFC2412705), National High Level Hospital Clinical Research Funding (2022-GSP-GG-18), Development Project of National Major Scientific Research Instrument (82327801) and Sanming Project of Medicine in Shenzhen (SZSM202011013).

Disclosures

The authors have no conflicts of interest to declare.

IRB Information

This retrospective study was approved by the Ethics Committee of Fuwai Hospital, Chinese Academy of Medical Sciences (No. 2023-1126).

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-23-0800

References

1. Lau W, Lee PT, Koh CH. Coronary artery anomalies - state of the art review. Curr Probl Cardiol 2023; 48: 101935.
2. Yun G, Nam TH, Chun EJ. Coronary artery fistulas: Pathophysiology, imaging findings, and management. Radiographics 2018; 38: 688–703.
3. El-Sabawi B, Al-Hijji MA, Eleid MF, Cabalka AK, Ammash NM, Dearani JA, et al. Transcatheter closure of coronary artery fistula: A 21-year experience. Catheter Cardiovasc Interv 2020; 96: 311–319.
4. Zhuang XF, Sun JP, Hou ZH, Li MZ, Sun XL. Left circumflex artery-left atria fistula treated with transcatheter closure. Circ Cardiovasc Imaging 2021; 14: e010913.
5. Pullela NK, Acharya D, Shanmugasundaram M, Acharya T, Ajmal M, Truong HT, et al. Transcatheter closure of complex left circumflex to coronary sinus fistula. JACC Case Rep 2022; 4: 559–563.
6. Al-Hijji M, El Sabbagh A, El Hajj S, AlKhouli M, El Sabawi B, Cabalka A, et al. Coronary artery fistulas: Indications, techniques, outcomes, and complications of transcatheter fistula closure. JACC Cardiovasc Interv 2021; 14: 1393–1406.
7. Zhu X, Zhou Q, Tong S, Zhou Y. Challenges and strategies in the management of coronary artery aneurysms. Hellenic J Cardiol 2021; 62: 112–120.
8. Podolec J, Wiewiórka Ł, Siudak Z, Malinowski K, Bartuś K, Dudek D, et al. Presence and characteristics of coronary artery fistulas among patients undergoing coronary angiography. Kardiol Pol 2019; 77: 1034–1039.
9. Li Y, Wei P, Xu Z, Xu L, Wan J, Zhang F, et al. Outcomes following transcatheter closure or surgical repair of coronary-left ventricular fistula: A single-center study with 10-year experience. Hellenic J Cardiol 2023; 74: 39–47.
10. Firouzi A, Alemzadeh-Ansari MJ, Mohebbi B, Khajali Z, Khalilipur E, Baay M, et al. Diverse transcatheter closure strategies in coronary artery fistulas a state-of-the-art approach. Curr Probl Cardiol 2022; 47: 101010.
11. Baumgartner H, De Backer J, Babu-Narayan SV, Budts W, Chessa M, Diller GP, et al. 2020 ESC guidelines for the management of adult congenital heart disease. Eur Heart J 2021; 42: 563–645.
12. Kalisz K, Sanders AE, Avery R, Allen BD. Coronary artery fistulas: A review of the current and future roles of imaging. J Thorac Imaging 2021; 36: 333–344.
13. Mangukia CV. Coronary artery fistula. Ann Thorac Surg 2012; 93: 2084–2092.
14. Hou B, Ma WG, Zhang J, Du M, Sun HS, Xu JP, et al. Surgical management of left circumflex coronary artery fistula: A 25-year single-center experience in 29 patients. Ann Thorac Surg 2014; 97: 530–536.
15. Armsby LR, Keane JF, Sherwood MC, Forbess JM, Perry SB, Lock JE. Management of coronary artery fistulae: Patient selection and results of transcatheter closure. J Am Coll Cardiol 2002; 39: 1026–1032.
16. Ilkay E, Celebi OO, Kacmaz F, Ozeke O. Percutaneous closure of coronary artery fistula: Long-term follow-up results. Postepy Kardiol Interwencyjnej 2015; 11: 318–322.
17. Abdelfattah OM, Saad AM, Kassis N, Shekhar S, Isogai T, Gad MM, et al. Utilization and outcomes of transcatheter coil embolization for various coronary artery lesions: Single-center 12-year experience. Catheter Cardiovasc Interv 2021; 98: 1317–1331.
18. Zhang ZG, Xu XD, Bai Y, Zhang XL, Tan HW, Zhu YF, et al. Transcatheter closure of medium and large congenital coronary artery fistula using wire-maintaining technique. J Cardiol 2015; 66: 509–513.
19. Buzaev IV, Plechev VV, Khalikova G, Khabirova K, Nikolaeva IE, Onorato EM. Case report: Coil occlusion of two congenital coronary cameral fistulas connecting right and left circumflex arteries to the right ventricle: An innovative stent-assisted technique. Front Cardiovasc Med 2021; 8: 769235.
20. Valsecchi O, Vassileva A, Martella F, Cereda AF. Trans-radial closure of coronary artery fistulae using Amplatzer vascular plug IV: A case series. Catheter Cardiovasc Interv 2019; 93: e394–e399.
21. Sumaya W, Horne D, Chen R, Elkhateeb O. Novel telescope anchoring technique for closure of gigantic coronary fistulae using 3-dimensional printing for planning. JACC Cardiovasc Interv 2022; 15: 896–897.
22. Latson LA. Coronary artery fistulas: How to manage them. Catheter Cardiovasc Interv 2007; 70: 110–116.
23. Jiang X, Liao S, Xiong B, Zheng L, Zhu H, Zhu X, et al. The coincidence of 3 different rare coronary artery anomalies in an adult patient with untreated Kawasaki disease: Case report. Medicine (Baltimore) 2016; 95: e3190.
24. Ouchi K, Sakuma T, Ojiri H. Coronary artery fistula in adults: Incidence and appearance on cardiac computed tomography and comparison of detectability and hemodynamic effects with those on transthoracic echocardiography. J Cardiol 2020; 76: 593–600.
25. Guglielmo M, Del Torto A, Muscogiuri G, Mostardini G, Pontone G. Multimodality imaging of a left circumflex artery to right atrium coronary artery fistula associated with giant aneurysm. Eur Heart J Cardiovasc Imaging 2021; 22: 20.
26. Gowda ST, Latson L, Sivakumar K, Hiremath G, Crystal M, Law M, et al. Anatomical classification and posttreatment remodeling characteristics to guide management and follow-up of neonates and infants with coronary artery fistula: A multicenter study from the Coronary Artery Fistula Registry. Circ Cardiovasc Interv 2021; 14: e009750.
27. Cao H, Qiu Y, Yuan D, Yu J, Li D, Jiang Y, et al. A computational fluid dynamics study pre- and post-fistula closure in a coronary artery fistula. Comput Methods Biomech Biomed Engin 2020; 23: 33–42.
28. Valente AM, Lock JE, Gauvreau K, Rodriguez-Huertas E, Joyce C, Armsby L, et al. Predictors of long-term adverse outcomes in patients with congenital coronary artery fistulae. Circ Cardiovasc Interv 2010; 3: 134–139.
29. Kemmochi R, Ohga Y, Kubo Y, Matsumoto M. Left circumflex artery aneurysm with fistula to the coronary sinus. Gen Thorac Cardiovasc Surg 2020; 68: 1196–1198.

Corresponding author

Register with J-STAGE for free!