Isolated congenital coronary artery fistula (CAF) is rare and varies with respect to size and hemodynamic significance. The prevalence of coronary artery ectasia in association with isolated congenital CAF, regardless of size, and after closure of large fistulae has not been systematically evaluated in the literature. This study aimed to characterize the demographic and echocardiographic differences between patients with large and small fistulae and to describe outcomes with respect to coronary ectasia in those who underwent closure. This is a retrospective review of an echocardiographic database that identified patients coded for CAF (1995 to 2012) and excluded those associated with complex cardiac disease and/or coronary anomalies and cardiomyopathy. Small fistulae were noted to arise mostly from the left anterior descending artery, drain into the pulmonary artery, and have a very low incidence of ectasia (n = 3 of 92), with a mean coronary artery diameter z score in these 3 patients of 3.45 ± 1.15. Larger fistulae had a female predominance, with most originating from the right coronary artery and draining into the right atrium; among the 12 patients who underwent procedural closure of large CAF, all feeding coronary arteries remained ectatic after closure, with a mean coronary artery diameter z score of 9.54 ± 5.66 after a total mean follow-up time of 3.95 ± 4.07 years. In conclusion, the occurrence of coronary dilatation justifies long-term follow-up irrespective of fistula size and successful closure of large CAF.
A coronary artery fistula (CAF) is an abnormal vascular communication between a coronary artery and cardiac chamber or major blood vessel. The overall incidence of CAF, although difficult to determine because many are small and incidentally discovered, is estimated at 0.002%. Angiography was considered the gold standard for evaluation of CAF, but echocardiography is a reliable noninvasive alternative. Clinical presentation depends on fistula size and ranges from being asymptomatic to exhibiting overt congestive heart failure. Small CAF are typically hemodynamically insignificant and frequently resolve spontaneously. Whereas observation with no intervention is generally recommended for smaller asymptomatic CAF, clinically significant fistulae normally require surgical or transcatheter closure (TCC), with comparable outcomes noted for both approaches. After CAF closure, the feeding coronary artery can undergo stenosis, thrombosis, or dilation. The prevalence of coronary ectasia in association with isolated congenital CAF, regardless of size, and after closure of large fistulae has not been systematically evaluated in the literature. The aims of this study were twofold: to characterize the demographic and echocardiographically determined morphologic differences between small and large CAF and to describe prevalence and persistence of coronary ectasia in both small and large fistulae.
Methods
This is a retrospective single-center study. Patients with congenital CAF were identified through an echocardiographic database search from the year 1995 to 2012. Institutional review board approval was obtained. Information related to patient demographics and presenting symptoms was collected through review of echocardiographic reports and medical records. Exclusion criteria included presence of significant congenital heart disease (e.g., hypoplastic left heart syndrome, pulmonary atresia with intact ventricular septum, tetralogy of Fallot, large ventricular septal defect, and so on), Kawasaki disease, presence of ventricular hypertrophy, evidence of acquired CAF, and/or presence of coronary anomalies (e.g., anomalous left coronary artery from the pulmonary artery).
Data from patients who were included in the study were then grouped and analyzed based on fistula size (determined qualitatively) and hemodynamic significance. In addition to other factors to assess for hemodynamic significance (e.g., left ventricular size), CAF were categorized in our laboratory as small if they were smaller in caliber compared with the feeding coronary and large if they were larger in caliber compared with the feeding coronary. Group A included patients with CAF initially described as small on echocardiographic reports. Group B consisted of subjects with medium-to-large, hemodynamically significant fistulae; this group also included some patients who had previously undergone surgical or interventional closure at outside institutions but obtained postclosure echocardiographic evaluations at our center.
Two-dimensional echocardiography was performed with commercially available standard equipment (Sequoia 256 [Siemens, Horten, Norway] and Philips iE33 [Philips Medical Systems, Andover, Massachusetts]) using transducers appropriate for patient age and size. CAF were identified by the presence of fistulous connections into a vessel or chamber by color flow mapping and Doppler interrogation. Attempts were made to identify both the coronary artery from which the fistula arose and the end point where the fistula emptied. The coronary artery caliber was typically measured in 2 dimensions from standard or modified parasternal short-axis views.
Demographic data pertaining to gender, age at echocardiographic diagnosis (or at first study at our center, if previously diagnosed or closed at outside institutions), years of follow-up, height, weight percentile, and body surface area at diagnosis or first echocardiography were collected.
If applicable, data pertaining to method of closure (i.e., through spontaneous resolution, surgery, or TCC), age at echocardiographically confirmed fistula closure, and presence or absence of coronary dilatation or ectasia (defined by proximal coronary artery diameter z score >+2.0 ) were noted. Among patients with follow-up data in group A, additional separate analyses focused on influence of demographics or echocardiographic features associated with spontaneous closure. In group B, trends in coronary artery dilation and regression were evaluated in the entire group inclusive of those with fistula closure and of those in whom the fistula was not closed at the time of the study.
Data from the 2 groups are summarized using percentages, means, SDs, medians, and ranges. Statistical analyses performed included Student and Welch t tests (when applicable) and Fisher’s exact test. A 2-tailed p value of ≤0.05 was considered to be statistically significant, whereas values >0.05 and ≤0.10 were considered to be trends.
Results
A total of 109 subjects with congenital CAF who underwent echocardiographic evaluation from the year 1995 to 2012 were included in the study. Group A comprised patients with fistulae identified as small ( Figure 1 ) and included 90 subjects with 92 fistulae; of these, 33 patients had follow-up studies and were included in further analysis focused on spontaneous resolution of small CAF. Group B comprised patients with medium-to-large, hemodynamically significant CAF ( Figure 1 ) and included 19 subjects with 20 fistulae. There was no significant difference in age at diagnosis (p = 0.14; Table 1 ). With respect to gender, there was a significantly greater predominance of female subjects in group B (p = 0.02; Table 2 ). Further information regarding demographics, fistula and feeding coronary artery characteristics, rate of spontaneous closure, significant differences between groups, and management is outlined in Table 1 and 2 . Presenting signs and symptoms are delineated in Table 3 . Of those who underwent procedural closure in group B, all patients demonstrated normal ventricular function on echocardiography (see later for median follow-up time).
| Demographic Information | Group A (n = 90) | Group B (n = 19) |
|---|---|---|
| Number of fistulae | 92 | 20 |
| Gender | ||
| Male | 57 (63%) | 6 (32%) |
| Female | 33 (37%) | 13 (68%) |
| Age at diagnosis (years) ∗ | ||
| Mean (SD) | 7.35 (8.59) | 15.73 (23.43) |
| Median (range) | 5.30 (0–66.30) | 4.00 (0–71.00) |
| Age at closure (years) † | ||
| Mean (SD) | 10.84 (6.93) | 23.19 (27.12) |
| Median (range) | 11.80 (0.40–26.30) | 9.20 (1.20–71.00) |
| Number of fistulae with spontaneous resolution | 16 (47%) ‡ | — |
| Number of fistulae closed procedurally | — | 13 |
| Surgical | — | 8 |
| Transcatheter | — | 5 |
| Follow-up time (years) | ||
| Mean (SD) | 1.26 (2.63) | 2.66 (3.64) |
| Median (range) | 0 (0–12.90) | 0.80 (0–11.40) |
∗ Also refers to age at initial echocardiography at our center if previously diagnosed elsewhere.
† Refers to patient age when closure is first documented on echocardiogram; applies to 15 group A patients with spontaneous resolution and all group B patients who underwent procedural closure for whom data are available.
‡ Refers to percentage of those for whom follow-up data are available (n = 34 fistulae).
| Characteristic | Group A (n = 90) | Group B (n = 19) | p Value |
|---|---|---|---|
| Female | 33 (37%) | 13 (68%) ∗ | 0.02 |
| Identified fistula origins | 16 fistulae (17%) | 20 fistulae (100%) ∗ | <0.01 |
| All left-sided | 15 fistulae (94%) ∗ | 10 fistulae (50%) | <0.01 |
| Left anterior descending artery | 7 fistulae (44%) ∗ | 2 fistulae (10%) | 0.05 |
| Right coronary artery | 1 fistula (6%) | 10 fistulae (50%) ∗ | <0.01 |
| Left main coronary artery | 2 fistulae (13%) | 3 fistulae (15%) | 1.00 |
| Circumflex artery | 1 fistula (6%) | 3 fistulae (15%) | 0.61 |
| Posterior descending artery | 2 fistulae (13%) | 0 fistulae (0%) | 0.19 |
| Identified fistula end points | 92 fistulae (100%) | 20 fistulae (100%) | 1.00 |
| Main pulmonary artery | 73 fistulae (79%) ∗ | 2 fistulae (10%) | <0.01 |
| Right pulmonary artery | 9 fistulae (10%) | 0 fistulae (0%) | 0.36 |
| Left pulmonary artery | 0 fistulae (0%) | 0 fistulae (0%) | 1.00 |
| Right ventricle | 6 fistulae (7%) | 6 fistulae (30%) ∗ | <0.01 |
| Left ventricle | 0 fistulae (0%) | 0 fistulae (0%) | 1.00 |
| Right atrium | 3 fistulae (3%) | 11 fistulae (55%) ∗ | <0.01 |
| Left atrium | 1 fistula (1%) | 1 fistula (5%) | 0.33 |
| Dilatation of feeding coronary artery | 3 patients (3%) | 19 patients (100%) ∗ | <0.01 |
| Signs or Symptoms | Group A (n = 90) | Group B (n = 19) |
|---|---|---|
| Murmur | 43 (48%) | 7 (37%) |
| Unknown or fistula diagnosis made at other center | 7 (8%) | 9 (47%) |
| Incidental finding on screening echocardiogram † | 34 (38%) | 1 (5%) |
| Arrhythmia/ECG abnormality | 10 (11%) | 1 (5%) |
| Chest pain | 7 (8%) | 2 (11%) |
| Pre-syncope/syncope | 4 (4%) | 0 |
| Palpitations | 3 (3%) | 2 (11%) |
| Cyanosis/desaturation | 2 (2%) | 0 |
| Heart failure ‡ | 0 | 2 (11%) |
| Dizziness | 1 (1%) | 0 |
| Fatigue | 1 (1%) | 0 |
| Prenatal diagnosis | 0 | 1 (5%) |
∗ Patients may present with >1 sign and/or symptom.
† For example, for such indications as prechemotherapy, genetic syndrome or connective tissue workup, evaluation of other suspected noncomplex congenital heart disease, and so on.
‡ Refers to heart failure secondary to hemodynamically significant coronary artery fistula.
In group A, the origins of most fistulae were unidentified; of those that were identified, the most common origin was the left anterior descending artery. In group B, most fistulae originated from the right coronary artery. All sites of termination were identified by echocardiography for both groups. The most common end point for group A CAF was the main pulmonary artery (MPA). The most common end point for group B fistulae was the right atrium ( Table 2 ).
Spontaneous resolution of CAF took place in group A fistulae only; closure of group B fistulae occurred either surgically (n = 8) or through TCC (n = 5). Of the 33 group A patients with follow-up (i.e., >1 echocardiographic study available), there was spontaneous closure of 16 of 34 fistulae (47%). Median follow-up time for these 33 patients was 2.20 years (range 0.10 to 12.90). When comparing the 16 fistulae in group A that underwent spontaneous resolution to the 18 that did not, there was no statistically significant difference between groups with respect to fistula origins, end points, gender, age at diagnosis, or association with coronary ectasia (p >0.10).
In group A, coronary artery ectasia (coronary artery z score >+2.0) was associated with small fistulae in only 3 cases; the mean coronary artery diameter z score in these 3 patients was 3.45 ± 1.15. Of all fistulae in the entire study, only 2 arose from the posterior descending artery (2%); left main coronary artery ectasia was associated with both of these small fistulae. The third fistula arose from the left anterior descending artery. End points included the right atrium, right ventricle, and MPA, respectively.
In group B, all fistulae were associated with coronary artery ectasia, regardless of whether they remained open or had undergone procedural closure. There were 9 patients with 10 fistulae who underwent closure at our center ( Table 4 and Figure 2 ). When comparing echocardiographic measurements taken before closure and at the final study after closure, there was no statistically significant difference in feeding coronary artery absolute size (p = 0.32), but a trend toward decrease in z score was noted (p = 0.08) with a median follow-up time for this subset of patients of 0.35 years (range 0.10 to 11.54).
| Statistical Function | Before Closure: Absolute Size (cm) | After Closure/Final Echocardiogram: Absolute Size (cm) | Before Closure: z-Score | After Closure/Final Echocardiogram: z-Score |
|---|---|---|---|---|
| Mean | 0.75 | 0.72 | 11.10 | 9.52 |
| Standard deviation | 0.35 | 0.38 | 6.21 | 6.13 |
| Median | 0.65 | 0.65 | 9.78 | 10.01 |
| Range | 0.41–1.61 | 0.31–1.61 | 4.39–22.20 | 2.43–20.85 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
