- Email: [email protected]
Arterial Switch Operation: Early and Late Outcome for Intramural Coronary Arteries
Arterial Switch Operation: Early and Late Outcome for Intramural Coronary Arteries Steven F. Thrupp, MA (Hons), MBChB, Thomas L. Gentles, FRACP, Alan R. Kerr, FRACS, and Kirsten Finucane, FRACS Paediatric and Congenital Cardiac Service, Starship Children’s Hospital, Auckland, New Zealand
PEDIATRIC CARDIAC
Background. Intramural coronary artery course (IMCA) is associated with an increased risk of coronary event and mortality after the arterial switch operation (ASO). We describe early and late outcomes at our institution from 1996 to 2006. Methods. Operation notes for all patients who underwent ASO within 60 days of birth were reviewed, and those with IMCA were identified. Mortality and morbidity were obtained from discharge summary, predischarge electrocardiogram, and echocardiogram. Follow-up included clinical review, electrocardiogram, echocardiography, dobutamine stress echocardiography, and angiography. Results. Eighteen patients of 215 in the cohort (8.4%) had IMCA. Intramural coronary artery course was more common in patients from French Polynesia (6 of 17; 35.3% versus 12 of 198; 6.1%; p ⴝ 0.001). Early mortality for
patients with IMCA was 1 of 18 (5.6%) compared with 6 of 197 (3%) for those without IMCA (p ⴝ 0.46). One IMCA patient was lost to follow-up. The remaining 16 are alive and asymptomatic. Of the 13 who underwent angiography, 2 had minor coronary artery stenoses at initial nonselective aortic root angiography. Both stenoses resolved at subsequent selective coronary angiography. None of the 11 who underwent dobutamine stress echocardiography had inducible ischemia. Conclusions. We report a high prevalence of IMCA in an ASO population, particularly among patients referred from French Polynesia. Intramural coronary artery course was not a risk factor for mortality after ASO. Angiography demonstrated excellent short-term and long-term structural outcome for IMCA. (Ann Thorac Surg 2012;94:2084 –90) © 2012 by The Society of Thoracic Surgeons
O
after ASO. Here we review our early and late outcomes for this population and compare mortality with our entire neonatal ASO cohort from 1996 to 2006.
verall postoperative mortality from the arterial switch operation (ASO) has been recently reported as between 1.1% [1] and 6.0% [2, 3]. Concern has persisted however as to what impact coronary artery anatomy may have on outcome, with intramural coronary artery morphology (IMCA) being identified as carrying significant additional risk of mortality [4 – 6] and coronary event [6 –10]. In 1994 Asou and colleagues [11] described a technique for detaching the posterior commissure of the aortic valve and unroofing the intramural segment of the coronary artery before transfer to the neoaorta using medially based trapdoor flaps. The commissure is then resuspended to the reconstructed neopulmonary artery. They reported 12 cases with no postoperative deaths, no ischemic changes on electrocardiogram (ECG), and no regional wall-motion abnormalities on a standard echocardiogram. We subsequently used this technique at our institution. Given the risks identified with coronary obstruction with IMCA we adopted an institutional policy of assessing these patients with dobutamine stress echocardiogram (DSE) and angiography at approximately 1 year Accepted for publication July 3, 2012. Address correspondence to Dr Finucane, Paediatric and Congenital Cardiac Service, Level 3, Starship Children’s Hospital, Private Bag 92024 (Park Rd, Grafton), Auckland, New Zealand; e-mail kfinucane@adhb. govt.nz.
© 2012 by The Society of Thoracic Surgeons Published by Elsevier Inc
Patients and Methods Patients This study was approved by our institutional review board, and individual patient consent was waived as individual patients are not identified. The operation reports of all 215 patients who received the ASO during the years 1996 to 2006 for d-transposition of the great arteries or double-outlet right ventricle within the first 60 days of life were reviewed, and the coronary artery pattern was identified in each case. Early death was defined as in-hospital or 30-day mortality. Data were obtained from our institution-wide mortality database. Cause of death was established from medical records or from post-mortem examination. Postoperative data obtained for the IMCA cohort included predischarge ECG, echocardiogram, and discharge summary. Late outcome was established from standard review by local pediatrician or pediatric cardiologist, most recent ECG, and most recent echocardiogram. Where possible, structural and functional outcomes were assessed by DSE and angiography at 12 months postoperatively and then subsequently as clinically indicated. 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2012.07.013
Ann Thorac Surg 2012;94:2084 –90
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
2085
Abbreviations and Acronyms ASO ⫽ arterial switch operation DSE ⫽ dobutamine stress echocardiogram ECG ⫽ electrocardiogram IMCA ⫽ intramural coronary artery pattern IQR ⫽ interquartile range LCA ⫽ left coronary artery RCA ⫽ right coronary artery
PEDIATRIC CARDIAC
Surgical Technique All but 2 patients underwent repair using the unroofing technique. Using this technique, once the heart is exposed and while the heart remains full, stay sutures are placed on the pulmonary artery at appropriate coronary artery transfer sites. After arterial and single or bicaval cannulation, hypothermia to a rectal temperature of 22°C Fig 2. Pericardial hood technique. (A) The upper edge is sutured to a scooped out edge, cut into the front of the neoaorta at the correct level. (B) An oval patch of pericardium is fashioned to cover the large button, creating a complete anterior pouch.
Fig 1. Coronary unroofing with transfer to medially based trapdoor flaps. (A) The aorta is divided well above the sinotubular junction. (B) Fine scissors are used to unroof the intramural segment of the coronary artery. (C) Each flap is elevated and excised cleanly using fine scissors or a blade. (D) The two coronaries are then excised with button tissue around each orifice and transferred, as usual, to medially based trapdoor flaps swung out from the front of the main pulmonary artery.
is induced. Single-dose cold-blood cardioplegia (45 to 60 ml/kg) is delivered over 5 minutes. Slush is placed at regular intervals. A left atrial vent is placed through a right atriotomy. The coronary branches are inspected and their position noted before cross-clamping. The ascending aorta is divided well above the sinotubular junction, and the coronary orifices are inspected from the aortic lumen. Some time is taken to ascertain the coronary anatomy. An incision into the lateral aspect of a sinus may allow better visualization of an intramural orifice. Gentle probing with a 1.5-mm coronary probe is performed to determine both the direction and length of the intramural pathway. The posterior commissure of the aortic valve is mobilized for 3 to 4 mm off the aortic wall. Fine scissors are used to open the roof of the intramural segment, and 6 – 0 Prolene (Ethicon, Somerville, NJ) is used to elevate the resulting flaps, which are then excised using either fine scissors or a blade. The aim is to unroof the intramural portion as far as possible while avoiding perforation through to the outside wall. This procedure increases the distance between the two coronary orifices. The two coronaries can then be excised with button tissue around each orifice and transferred to medially based trapdoor flaps swung out from the front of the main pulmonary artery (Fig 1). The posterior commissure is then resuspended to an autologous pericardial patch as the neo–main pulmonary artery is reconstructed. When there is a single orifice with one or more intramural branches, the coronary button is large and cannot be turned the 90 degrees necessary for the trapdoor technique. In this case the button is left with the orifices facing forward. The upper edge of the button is sutured to a scooped out edge cut into the front of the neoaorta at the correct level, and an oval patch of pericardium is fashioned to cover it over and create a complete anterior
2086
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
Ann Thorac Surg 2012;94:2084 –90
Table 1. Coronary Artery Patternsa Found in Arterial Switch Operation Population 1996 –2006 Coronary Artery Pattern
PEDIATRIC CARDIAC
Most common pattern Second most common pattern Intramural coronary artery Inverted circumflex Single orifice Inverted Single sinus Other coronary pattern Total
Number
%
136 25 18 15 10 6 3 2 215
63.3 11.6 8.4 7.0 4.7 2.8 1.4 0.9 100
a The most common pattern observed is left coronary artery (LCA) arising from the left-facing sinus (LFS) before dividing into left anterior descending artery (LAD) and circumflex artery; right coronary artery (RCA) arises from right-facing sinus (RFS). The second most common pattern observed is the LCA and RCA arising from their appropriate facing sinuses, but with the circumflex artery arising from the RCA. The intramural coronary artery pattern is observed when one or both coronary arteries have an intramural origin. The inverted circumflex pattern is observed when the RCA arises from the LCA, and the circumflex arises from the RFS. A single orifice is the pattern observed when all coronary arteries arise from a single orifice. An inverted condition occurs when the RCA arises from the LFS, and the LCA arises from the RFS. A single sinus pattern is observed when two orifices arise from a single sinus. Anything other than these defined patterns is termed other coronary pattern.
pouch (Fig 2). With this technique there is potential for coronary compression from the pulmonary artery, particularly when pulmonary artery pressure is elevated. It is most important that the neopulmonary artery and its confluence straddle the neoaorta superior to the coronary pouch and not directly over it. For this reason the ascending aorta must be cannulated distally and divided well above the coronary orifices. The anterior pouch technique was used twice in our series. In one case this was for the indication given above. In the second case it was used in a patient with the most
common IMCA pattern (see below) during the period before our use of the coronary unroofing technique.
Statistical Analysis Comparison between patient groups was made with 2 test or Fisher’s exact test (for categorical variables) and nonparametric Mann-Whitney U test (for continuous variables). Categorical variables are expressed as frequency and percentage. Continuous variables are expressed as median and interquartile range (IQR). Timerelated survival data were analyzed using Kaplan-Meier analysis. Data analysis was performed using SAS statistical package, version 9.1.3 (SAS Institute, Cary, NC). A two-sided probability value of less than 0.05 was considered to be significant.
Results Anatomic Observations The prevalence of coronary artery patterns for all 215 patients is described in Table 1. Eighteen patients (8.4%) were found to have an IMCA pattern. Their demographic, anatomic, and perioperative characteristics are summarized and compared with the non-IMCA population (Table 2). The specific morphology of the various IMCA patterns was identified. These were classified using five anatomic variations originally identified by Asou and colleagues [11] and diagrammatically represented in that report. Of the 18 patients with IMCA, 11 had an intramural pattern of both main coronary arteries originating from the right-facing sinus through separate orifices, so that the left coronary artery (LCA) was intramural, passing behind the posterior commissure before exiting from the left-facing sinus and dividing to a left anterior descend-
Table 2. Intraoperative and Early Postoperative Characteristics of Intramural Coronary Artery Pattern and Non–Intramural Coronary Artery Pattern Populationa Characteristics Age at operation (days) Weight at operation (kg) Sex (male) VSD Side-by-side great arteries Hypoplastic arch/coarctation Bypass time (min) Cross-clamp time (min) Postoperative ICU stay (days) Postoperative length of stay (days) Delayed sternal closure Mortality a
IMCA (n ⫽ 18)
Non-IMCA (n ⫽ 197)
p Value
9 (8, 11) 3.6 (3.4, 3.8) 14 (77.8) 6 (33.3) 0 (0) 2 (11.1) 159.5 (141, 186) 86.5 (80, 108) 3.0 (2.5, 4.5) 7.5 (7, 10) 5 (27.8) 1 (5.6)
8 (6, 11) 3.5 (3.0, 3.9) 140 (71.1) 59 (30.0) 16 (8.1) 8 (4.1) 150 (134, 185.5) 78 (71, 102) 3.6 (2.6, 6.4) 11 (8, 14) 32 (16.2) 6 (3.0)
0.37 0.28 0.55 0.76 0.37 0.20 0.33 0.02 0.08 0.0029 0.21 0.46
Continuous data are expressed as median (interquartile range) and categorical data as frequency (percentage).
ICU ⫽ intensive care unit;
IMCA ⫽ intramural coronary artery pattern;
VSD ⫽ ventricular septal defect.
Ann Thorac Surg 2012;94:2084 –90
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
2087
Table 3. Mortality: Preoperative, Operative, and Anatomic Features of Those Patients Who Died Compared With Survivorsa Mortality
Age at operation (days) Weight at operation (kg) Sex (male) Surgical era of ASO 1996–2001 2002–2006 Coronary artery patternb Intramural Inverted Most common pattern Other coronary pattern Inverted circumflex Second most common pattern Single orifice Single sinus Side-by-side great arteries Ventricular septal defect Hypoplastic arch/coarctation Complex TGA Bypass time (min) Cross-clamp time (min) Delayed sternal closure
No (n ⫽ 208)
Yes (n ⫽ 7)
8 (6, 11) 3.5 (3.1, 3.9) 147 (70.7)
28 (11, 44) 3.0 (2.4, 4.1) 7 (100)
122 (58.7) 86 (41.3)
0.0014 0.51 0.20 0.70
5 (71.4) 2 (28.6) 0.74
17 (8.2) 6 (2.9) 131 (63.0) 2 (1.0) 14 (6.7) 25 (12.0) 10 (4.8) 3 (1.4) 14 (6.7) 61 (29.3) 9 (4.3) 12 (6.1) 150 (134, 182) 78.5 (72, 101) 32 (15.4)
1 (14.3) 0 (0) 5 (71.4) 0 (0) 1 (14.3) 0 (0) 0 (0) 0 (0) 2 (28.6) 4 (57.1) 1 (14.3) 5 (71.4%) 351 (197, 391) 185 (93, 232) 5 (71.4)
a Continuous data are expressed as median (interquartile range) and categorical data as frequency (percentage). coronary artery patterns, see Table 1.
ASO ⫽ arterial switch operation;
p Value
0.09 0.20 0.29 0.0001 0.0003 0.0021 0.0019 b
For definitions of the various
TGA ⫽ transposition of the great arteries.
ing (LAD) and circumflex coronary artery, while the right coronary artery (RCA) was normal (variation 1a). Three other patients had close variations on this pattern. One patient had a single orifice (variation 1b), whereas another had separate origins from the rightfacing sinus and the circumflex arose from the RCA while the left anterior descending coronary artery ran an intramural course (variation 2). In the third patient, the origin of the intramural LCA was from above the posterior commissure (variation 3). The remaining 4 patients had different IMCA patterns. Two patients had a normally arising LCA but their RCA had a slitlike orifice at or above the sinotubular junction that traveled intramurally down the right-facing sinus before exiting normally onto the surface of the heart (variation 4). Two patients had bilateral IMCA patterns (variation 5). The first of these had bilateral origins above the sinotubular junction of the appropriate sinus that ran intramurally before exiting normally. The second of these had a single orifice well up the aorta to the right of the posterior commissure that immediately bifurcated to an intramural RCA that exited on the rightfacing sinus before giving off a circumflex branch, and an intramural left anterior descending coronary artery that took a long intramural course before exiting on the left-facing sinus.
Demographic Observations Of the 215 patients, 17 were referred from French Polynesia. Of these 17, 6 had IMCA, giving an overall prevalence of 35.3% in this population compared with 6.1% (12 of 198) in the remaining non–French Polynesian population (p ⫽ 0.001).
Early Outcome: Mortality There were 7 early deaths in our entire ASO cohort, giving an overall mortality of 3.26% (Table 3). There was 1 death in the IMCA population (5.56%) and 6 in the non-IMCA population (3.0%; p ⫽ 0.46). No single anatomic feature was associated with early mortality. However, when patients were grouped according to the presence or absence of concomitant cardiac comorbidity, statistical significance was reached. Concomitant cardiac comorbidity was defined as the presence of Taussig-Bing anomaly, or any aortic, pulmonary, or outflow tract abnormality requiring intervention, or low operative weight (1 kg or less). The additional complexities associated with these comorbidities are reflected in significantly longer cardiopulmonary bypass time, longer cross-clamp time, and a greater prevalence of unstable postoperative patients evident by delayed sternal closure. The older age at operation in the 7 who died reflects 3 premature neonates
PEDIATRIC CARDIAC
Characteristics
2088
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
Ann Thorac Surg 2012;94:2084 –90
Fig 3. Kaplan-Meier survival probability curves for intramural coronary artery pattern (IMCA) and non-IMCA cohorts.
PEDIATRIC CARDIAC
in whom treatment was delayed, and 2 others requiring transfer from Pacific Islands to New Zealand for treatment, both of whom died of sepsis attributable to Pseudomonas species. The neonate with the IMCA pattern who died had a Taussig-Bing type malformation. This neonate was transferred ventilated from Tahiti and underwent ASO at day 8 of life. Thirty-six hours postoperatively there was low cardiac output and leukopenia. The neonate was treated for sepsis but died several hours later. Post-mortem examination confirmed disseminated Pseudomonas infection.
Early Outcome: Morbidity (Intramural Coronary Artery Pattern) Intraoperatively, 1 of the 18 IMCA patients had ischemic changes on ECG when weaning from bypass. This patient was the first in our series and underwent repair using the pericardial hood technique. The ischemia was attributable to dynamic compression of the pericardial hood by the neopulmonary artery. This was resolved by elongation of the neopulmonary artery, plication of the posterior pericardial patch of the neoaorta, and the placement of anterior plication sutures on the aorta to change the angulation of the pericardial hood. Postoperatively, sternal closure was delayed in 5 patients. Two patients exhibited tachyarrhythmias that required chemical or electrical cardioversion. One patient experienced a focal seizure with no neurologic sequelae. Two patients had preoperative comorbidities that complicated their postoperative course. One of these had necrotizing enterocolitis. The other had an embolic cerebral infarct from femoral vein thrombosis after repeated femoral vein instrumentation before ASO. ELECTROCARDIOGRAM. One patient had ST-segment elevation
on ECG in the hours after surgery that subsequently resolved. There were no ischemic changes on ECG before discharge for all 17 surviving patients. All patients were in sinus rhythm with or without right bundle-branch block.
Before hospital discharge no patient had significant left or right ventricular dysfunction. One had mild neoaortic regurgitation; 4 had mild neopulmonary regurgitation. There were no residual atrial or ventricular septal defects found that required subsequent closure.
ECHOCARDIOGRAM.
Late Outcome: Mortality Follow-up for the entire cohort of 215 patients is 9.1 years (IQR, 6.0 to 11.4). There was no significant difference in survival. Intramural coronary artery pattern survival was 94.4% at 1, 5, and 10 years compared with 95.9%, 94.9%, and 94.3% during the same interval for the non-IMCA population (p ⫽ 0.99; Fig 3).
Late Outcome: Morbidity (Intramural Coronary Artery Pattern) After discharge each patient with IMCA was clinically assessed by a pediatrician or pediatric cardiologist and received ECG, echocardiogram, angiogram, and DSE. In some cases not all of these investigations were possible. Of the 17 surviving cases, 1 was lost to follow-up and only received an echocardiogram during an outpatient visit immediately after discharge. Of the remaining 16, 2 were from remote Pacific islands and did not return for catheterization or DSE as transfer was not considered justified in the context of being asymptomatic with normal ECG and echocardiogram. Another patient did not receive angiography or DSE because of unrelated illness. Two patients received angiography (which was normal) but did not receive DSE for reasons that are not documented. CLINICAL STATUS. Of the 16 patients who had long-term follow-up, all are asymptomatic in terms of their cardiovascular health (median time of follow up, 81.5 months; IQR, 50.5 to 103 months). ELECTROCARDIOGRAM. An ECG was available in 16 of 17 patients. There were no ischemic changes found at a
median period of 61 months (IQR, 37 to 77 months) after surgery. ECHOCARDIOGRAM. Sixteen patients received outpatient echocardiograms at a median period of 41 months (IQR, 23 to 77 months) after surgery. Left and right ventricular function was normal in all studies. Three patients exhibited neoaortic root dilation and mild or moderate aortic valve regurgitation, and 2 had mild neopulmonary regurgitation.
Nonselective aortic root angiography was undertaken in 13 patients at a median time of 13 months (IQR, 13 to 22 months) postoperatively. Coronary artery abnormalities warranting subsequent angiography were found in 2 patients. In 1 patient moderate left main stem stenosis was demonstrated at 22 months. A selective coronary angiogram at 91 months showed diffuse narrowing but no significant stenosis. The initial nonselective angiogram of the second patient occurred at 12 months postoperatively and showed mild stenosis of the LCA. Selective coronary angiogram at 42 months demonstrated a widely patent left coronary ostium.
ANGIOGRAPHY.
Eleven patients received DSE at a median time of 13 months (IQR, 12.5 to 28 months). All studies were negative for inducible ischemia. Two patients who did not receive DSE had normal angiography.
DOBUTAMINE STRESS ECHOCARDIOGRAM.
CORONARY INTERVENTIONS IN THE OVERALL ARTERIAL SWITCH OPERATION POPULATION. There have been no coronary interventions in the IMCA cohort to date and one intervention in the non-IMCA cohort (p ⫽ 1.0). This was in a child with the most common pattern who underwent ASO in 1996, and who had an internal mammary artery graft 8.5 years later.
Comment The prevalence of IMCA in this study is higher than previously noted in other reports [6, 11–14], including from our own institution [15]. In particular we have identified a subset of patients from French Polynesia with a high prevalence. The difference in prevalence between our current and previous reports is owing to the previous study describing an earlier period of time from 1984 to 1999 during which successful transfers from French Polynesia were less frequent. In addition, the current report focuses solely on a neonatal population. The reasons for the increased prevalence of IMCA in the French Polynesian cohort are not clear and are beyond the scope of the current paper; however, the finding raises the possibility of a genetic basis or alternatively a common environmental exposure. The predominant pattern we noted of an intramural LCA originating on the right-facing sinus and passing behind the posterior aortic commissure is consistent with
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
2089
that reported in a previous large series [6] and composed a significant number of the intramural cases to which Asou and colleagues originally applied the unroofing technique [11]. The presence of an intramural coronary artery did not carry an additional risk of early death in our population as has been found in other studies [5, 6, 15]. In fact, no single anatomic feature was found to be significantly associated with early mortality. However, additional cardiac comorbidities were significantly associated with early death. The additional operative burden associated with these comorbidities was reflected in significantly prolonged bypass and aortic cross-clamp times, and is an association seen in other studies [4, 15, 16]. The older age at operation seen in those who died is attributable to the prevalence of premature neonates in this group who had prolonged medical treatment before surgery, and to 2 neonates who were transferred from remote Pacific islands. Others have found that intramural coronary arteries are associated with coronary stenosis and coronary event after ASO [6, 8 –10]. For this reason we followed our patients with nonselective angiography and DSE routinely at 1 year after ASO. Our results showed few cases of coronary artery stenosis using the unroofing technique. Although we do not routinely follow up our general cohort with the same degree of structural evaluation unless clinically indicated, we note comparable and low rates of coronary reintervention procedures in both groups. Where stenosis was suspected in the IMCA group at 12 months by nonselective angiography, subsequent selective angiography demonstrated reduction in stenosis with time. This finding could be consistent with a remodeling process, or could reflect the inaccuracy of nonselective coronary imaging in infants. Given that angiography itself is not without risk, we now consider that 12-month follow-up angiography be deferred in patients with a negative DSE until an age when selective coronary angiography or high-resolution computed tomography is possible. In view of recent results with high-resolution computed tomography our current approach is to image noninvasively and then to follow with conventional selective coronary angiography if clinically warranted [17, 18].
Study Limitations Given our sample size it would take 3 deaths in the IMCA cohort (giving a mortality of 16.7%) compared with a mortality of 3% in the non-IMCA cohort for significance to be reached. The lack of significant difference in mortality we found may therefore be related to limited statistical power given the sample size.
Conclusions We report outcomes for a cohort of patients with IMCA who underwent ASO and who predominantly received coronary unroofing repair. There was a high prevalence of IMCA in this population (8.4%), especially in a subset of patients referred from Tahiti (35.3%). There was no
PEDIATRIC CARDIAC
Ann Thorac Surg 2012;94:2084 –90
2090
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
significant difference in mortality between IMCA and non-IMCA patients. Patients are asymptomatic at a median follow-up time of 81.5 months with excellent structural and functional outcome as assessed by angiography and DSE. We have obtained acceptable early and late outcomes for patients with d-transposition of the great arteries or double-outlet right ventricle and an IMCA using the coronary unroofing technique.
PEDIATRIC CARDIAC
The authors gratefully acknowledge the contribution of Steve Withy, Mildred Lee and Felicity Roberts for their respective assistance with data acquisition, statistical analysis and manuscript preparation.
References 1. Freed DH, Robertson CMT, Sauve RS, et al. Intermediateterm outcomes of the arterial switch operation for transposition of great arteries in neonates: alive but well? J Thorac Cardiovasc Surg 2006;132:845–52. 2. Sarris GE, Chatzis AC, Giannopoulos NM, et al. The arterial switch operation in Europe for transposition of the great arteries: a multi-institutional study from the European Congenital Heart Surgeons Association. J Thorac Cardiovasc Surg 2006;132:633–9. 3. Armishaw J, Gentles TL, Calder AL, Raudkivi PJ, Kerr AR. Transposition of the great arteries: operative outcome in the current era. N Z Med J 2000;113:456 –9. 4. Kirklin JW, Blackstone EH, Tchervenkov CI, Castaneda AR. Clinical outcomes after the arterial switch operation for transposition. Patient, support, procedural and institutional risk factors. Congenital Heart Surgeons Society. Circulation 1992;86:1501–15. 5. Pasquali SK, Hasselblad V, Ji JS, Kong DF, Sanders SP. Coronary artery pattern and outcome of arterial switch operation for transposition of the great arteries: a metaanalysis. Circulation 2002;106:2575– 80. 6. Metton O, Calvaruso D, Gaudin R, et al. Intramural coronary arteries and outcome of neonatal arterial switch operation. Eur J Cardiothorac Surg 2010;37:1246 –53.
Ann Thorac Surg 2012;94:2084 –90
7. Legendre A, Losay J, Touchot-Koné A, et al. Coronary events after arterial switch operation for transposition of the great arteries. Circulation 2003;108(Suppl 1):II-186 –90. 8. Prêtre R, Tamisier D, Bonhoeffer P, et al. Results of the arterial switch operation in neonates with transposed great arteries. Lancet 2001;357:1826 –30. 9. Tamisier D, Ouaknine R, Pouard P, et al. Neonatal arterial switch operation: coronary artery patterns and coronary events. Eur J Cardiothorac Surg 1997;11:810 –7. 10. Bonhoeffer P, Bonnet D, Piéchaud J-F, et al. Coronary artery obstruction after the arterial switch operation for transposition of the great arteries in newborns. J Am Coll Cardiol 1997;29:202– 6. 11. Asou T, Karl TR, Pawade A, Mee RB. Arterial switch: translocation of the intramural coronary artery. Ann Thorac Surg 1994;57:461–5. 12. Hutter PA, Bennink GBWE, Ay L, Raes IB, Hitchcock JF, Meijboom EJ. Influence of coronary anatomy and reimplantation on the long-term outcome of the arterial switch. Eur J Cardiothorac Surg 2000;18:207–13. 13. Sachweh JS, Tiete AR, Jockenhoevel S, et al. Fate of intramural coronary arteries after arterial switch operation. Thorac Cardiovasc Surg 2002;50:40 – 4. 14. Qamar ZA, Goldberg CS, Devaney EJ, Bove EL, Ohye RG. Current risk factors and outcomes for the arterial switch operation. Ann Thorac Surg 2007;84:871–9. 15. Wong SH, Finucane K, Kerr AR, O’Donnell C, West T, Gentles TL. Cardiac outcome up to 15 years after the arterial switch operation. Heart Lung Circ 2008;17:48 –53. 16. Blume ED, Altmann K, Mayer JE, Colan SD, Gauvreau K, Geva T. Evolution of risk factors influencing early mortality of the arterial switch operation. J Am Coll Cardiol 1999;33: 1702–9. 17. Ou P, Mousseaux E, Azarine A, et al. Detection of coronary complications after the arterial switch operation for transposition of the great arteries: first experience with multislice computed tomography in children. J Thorac Cardiovasc Surg 2006;131:639 – 43. 18. Ou P, Celermajer DS, Marini D, et al. Safety and accuracy of 64-slice computed tomography coronary angiography in children after the arterial switch operation for transposition of the great arteries. JACC Cardiovasc Imaging 2008;1:331–9.
PEDIATRIC CARDIAC
Background. Intramural coronary artery course (IMCA) is associated with an increased risk of coronary event and mortality after the arterial switch operation (ASO). We describe early and late outcomes at our institution from 1996 to 2006. Methods. Operation notes for all patients who underwent ASO within 60 days of birth were reviewed, and those with IMCA were identified. Mortality and morbidity were obtained from discharge summary, predischarge electrocardiogram, and echocardiogram. Follow-up included clinical review, electrocardiogram, echocardiography, dobutamine stress echocardiography, and angiography. Results. Eighteen patients of 215 in the cohort (8.4%) had IMCA. Intramural coronary artery course was more common in patients from French Polynesia (6 of 17; 35.3% versus 12 of 198; 6.1%; p ⴝ 0.001). Early mortality for
patients with IMCA was 1 of 18 (5.6%) compared with 6 of 197 (3%) for those without IMCA (p ⴝ 0.46). One IMCA patient was lost to follow-up. The remaining 16 are alive and asymptomatic. Of the 13 who underwent angiography, 2 had minor coronary artery stenoses at initial nonselective aortic root angiography. Both stenoses resolved at subsequent selective coronary angiography. None of the 11 who underwent dobutamine stress echocardiography had inducible ischemia. Conclusions. We report a high prevalence of IMCA in an ASO population, particularly among patients referred from French Polynesia. Intramural coronary artery course was not a risk factor for mortality after ASO. Angiography demonstrated excellent short-term and long-term structural outcome for IMCA. (Ann Thorac Surg 2012;94:2084 –90) © 2012 by The Society of Thoracic Surgeons
O
after ASO. Here we review our early and late outcomes for this population and compare mortality with our entire neonatal ASO cohort from 1996 to 2006.
verall postoperative mortality from the arterial switch operation (ASO) has been recently reported as between 1.1% [1] and 6.0% [2, 3]. Concern has persisted however as to what impact coronary artery anatomy may have on outcome, with intramural coronary artery morphology (IMCA) being identified as carrying significant additional risk of mortality [4 – 6] and coronary event [6 –10]. In 1994 Asou and colleagues [11] described a technique for detaching the posterior commissure of the aortic valve and unroofing the intramural segment of the coronary artery before transfer to the neoaorta using medially based trapdoor flaps. The commissure is then resuspended to the reconstructed neopulmonary artery. They reported 12 cases with no postoperative deaths, no ischemic changes on electrocardiogram (ECG), and no regional wall-motion abnormalities on a standard echocardiogram. We subsequently used this technique at our institution. Given the risks identified with coronary obstruction with IMCA we adopted an institutional policy of assessing these patients with dobutamine stress echocardiogram (DSE) and angiography at approximately 1 year Accepted for publication July 3, 2012. Address correspondence to Dr Finucane, Paediatric and Congenital Cardiac Service, Level 3, Starship Children’s Hospital, Private Bag 92024 (Park Rd, Grafton), Auckland, New Zealand; e-mail kfinucane@adhb. govt.nz.
© 2012 by The Society of Thoracic Surgeons Published by Elsevier Inc
Patients and Methods Patients This study was approved by our institutional review board, and individual patient consent was waived as individual patients are not identified. The operation reports of all 215 patients who received the ASO during the years 1996 to 2006 for d-transposition of the great arteries or double-outlet right ventricle within the first 60 days of life were reviewed, and the coronary artery pattern was identified in each case. Early death was defined as in-hospital or 30-day mortality. Data were obtained from our institution-wide mortality database. Cause of death was established from medical records or from post-mortem examination. Postoperative data obtained for the IMCA cohort included predischarge ECG, echocardiogram, and discharge summary. Late outcome was established from standard review by local pediatrician or pediatric cardiologist, most recent ECG, and most recent echocardiogram. Where possible, structural and functional outcomes were assessed by DSE and angiography at 12 months postoperatively and then subsequently as clinically indicated. 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2012.07.013
Ann Thorac Surg 2012;94:2084 –90
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
2085
Abbreviations and Acronyms ASO ⫽ arterial switch operation DSE ⫽ dobutamine stress echocardiogram ECG ⫽ electrocardiogram IMCA ⫽ intramural coronary artery pattern IQR ⫽ interquartile range LCA ⫽ left coronary artery RCA ⫽ right coronary artery
PEDIATRIC CARDIAC
Surgical Technique All but 2 patients underwent repair using the unroofing technique. Using this technique, once the heart is exposed and while the heart remains full, stay sutures are placed on the pulmonary artery at appropriate coronary artery transfer sites. After arterial and single or bicaval cannulation, hypothermia to a rectal temperature of 22°C Fig 2. Pericardial hood technique. (A) The upper edge is sutured to a scooped out edge, cut into the front of the neoaorta at the correct level. (B) An oval patch of pericardium is fashioned to cover the large button, creating a complete anterior pouch.
Fig 1. Coronary unroofing with transfer to medially based trapdoor flaps. (A) The aorta is divided well above the sinotubular junction. (B) Fine scissors are used to unroof the intramural segment of the coronary artery. (C) Each flap is elevated and excised cleanly using fine scissors or a blade. (D) The two coronaries are then excised with button tissue around each orifice and transferred, as usual, to medially based trapdoor flaps swung out from the front of the main pulmonary artery.
is induced. Single-dose cold-blood cardioplegia (45 to 60 ml/kg) is delivered over 5 minutes. Slush is placed at regular intervals. A left atrial vent is placed through a right atriotomy. The coronary branches are inspected and their position noted before cross-clamping. The ascending aorta is divided well above the sinotubular junction, and the coronary orifices are inspected from the aortic lumen. Some time is taken to ascertain the coronary anatomy. An incision into the lateral aspect of a sinus may allow better visualization of an intramural orifice. Gentle probing with a 1.5-mm coronary probe is performed to determine both the direction and length of the intramural pathway. The posterior commissure of the aortic valve is mobilized for 3 to 4 mm off the aortic wall. Fine scissors are used to open the roof of the intramural segment, and 6 – 0 Prolene (Ethicon, Somerville, NJ) is used to elevate the resulting flaps, which are then excised using either fine scissors or a blade. The aim is to unroof the intramural portion as far as possible while avoiding perforation through to the outside wall. This procedure increases the distance between the two coronary orifices. The two coronaries can then be excised with button tissue around each orifice and transferred to medially based trapdoor flaps swung out from the front of the main pulmonary artery (Fig 1). The posterior commissure is then resuspended to an autologous pericardial patch as the neo–main pulmonary artery is reconstructed. When there is a single orifice with one or more intramural branches, the coronary button is large and cannot be turned the 90 degrees necessary for the trapdoor technique. In this case the button is left with the orifices facing forward. The upper edge of the button is sutured to a scooped out edge cut into the front of the neoaorta at the correct level, and an oval patch of pericardium is fashioned to cover it over and create a complete anterior
2086
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
Ann Thorac Surg 2012;94:2084 –90
Table 1. Coronary Artery Patternsa Found in Arterial Switch Operation Population 1996 –2006 Coronary Artery Pattern
PEDIATRIC CARDIAC
Most common pattern Second most common pattern Intramural coronary artery Inverted circumflex Single orifice Inverted Single sinus Other coronary pattern Total
Number
%
136 25 18 15 10 6 3 2 215
63.3 11.6 8.4 7.0 4.7 2.8 1.4 0.9 100
a The most common pattern observed is left coronary artery (LCA) arising from the left-facing sinus (LFS) before dividing into left anterior descending artery (LAD) and circumflex artery; right coronary artery (RCA) arises from right-facing sinus (RFS). The second most common pattern observed is the LCA and RCA arising from their appropriate facing sinuses, but with the circumflex artery arising from the RCA. The intramural coronary artery pattern is observed when one or both coronary arteries have an intramural origin. The inverted circumflex pattern is observed when the RCA arises from the LCA, and the circumflex arises from the RFS. A single orifice is the pattern observed when all coronary arteries arise from a single orifice. An inverted condition occurs when the RCA arises from the LFS, and the LCA arises from the RFS. A single sinus pattern is observed when two orifices arise from a single sinus. Anything other than these defined patterns is termed other coronary pattern.
pouch (Fig 2). With this technique there is potential for coronary compression from the pulmonary artery, particularly when pulmonary artery pressure is elevated. It is most important that the neopulmonary artery and its confluence straddle the neoaorta superior to the coronary pouch and not directly over it. For this reason the ascending aorta must be cannulated distally and divided well above the coronary orifices. The anterior pouch technique was used twice in our series. In one case this was for the indication given above. In the second case it was used in a patient with the most
common IMCA pattern (see below) during the period before our use of the coronary unroofing technique.
Statistical Analysis Comparison between patient groups was made with 2 test or Fisher’s exact test (for categorical variables) and nonparametric Mann-Whitney U test (for continuous variables). Categorical variables are expressed as frequency and percentage. Continuous variables are expressed as median and interquartile range (IQR). Timerelated survival data were analyzed using Kaplan-Meier analysis. Data analysis was performed using SAS statistical package, version 9.1.3 (SAS Institute, Cary, NC). A two-sided probability value of less than 0.05 was considered to be significant.
Results Anatomic Observations The prevalence of coronary artery patterns for all 215 patients is described in Table 1. Eighteen patients (8.4%) were found to have an IMCA pattern. Their demographic, anatomic, and perioperative characteristics are summarized and compared with the non-IMCA population (Table 2). The specific morphology of the various IMCA patterns was identified. These were classified using five anatomic variations originally identified by Asou and colleagues [11] and diagrammatically represented in that report. Of the 18 patients with IMCA, 11 had an intramural pattern of both main coronary arteries originating from the right-facing sinus through separate orifices, so that the left coronary artery (LCA) was intramural, passing behind the posterior commissure before exiting from the left-facing sinus and dividing to a left anterior descend-
Table 2. Intraoperative and Early Postoperative Characteristics of Intramural Coronary Artery Pattern and Non–Intramural Coronary Artery Pattern Populationa Characteristics Age at operation (days) Weight at operation (kg) Sex (male) VSD Side-by-side great arteries Hypoplastic arch/coarctation Bypass time (min) Cross-clamp time (min) Postoperative ICU stay (days) Postoperative length of stay (days) Delayed sternal closure Mortality a
IMCA (n ⫽ 18)
Non-IMCA (n ⫽ 197)
p Value
9 (8, 11) 3.6 (3.4, 3.8) 14 (77.8) 6 (33.3) 0 (0) 2 (11.1) 159.5 (141, 186) 86.5 (80, 108) 3.0 (2.5, 4.5) 7.5 (7, 10) 5 (27.8) 1 (5.6)
8 (6, 11) 3.5 (3.0, 3.9) 140 (71.1) 59 (30.0) 16 (8.1) 8 (4.1) 150 (134, 185.5) 78 (71, 102) 3.6 (2.6, 6.4) 11 (8, 14) 32 (16.2) 6 (3.0)
0.37 0.28 0.55 0.76 0.37 0.20 0.33 0.02 0.08 0.0029 0.21 0.46
Continuous data are expressed as median (interquartile range) and categorical data as frequency (percentage).
ICU ⫽ intensive care unit;
IMCA ⫽ intramural coronary artery pattern;
VSD ⫽ ventricular septal defect.
Ann Thorac Surg 2012;94:2084 –90
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
2087
Table 3. Mortality: Preoperative, Operative, and Anatomic Features of Those Patients Who Died Compared With Survivorsa Mortality
Age at operation (days) Weight at operation (kg) Sex (male) Surgical era of ASO 1996–2001 2002–2006 Coronary artery patternb Intramural Inverted Most common pattern Other coronary pattern Inverted circumflex Second most common pattern Single orifice Single sinus Side-by-side great arteries Ventricular septal defect Hypoplastic arch/coarctation Complex TGA Bypass time (min) Cross-clamp time (min) Delayed sternal closure
No (n ⫽ 208)
Yes (n ⫽ 7)
8 (6, 11) 3.5 (3.1, 3.9) 147 (70.7)
28 (11, 44) 3.0 (2.4, 4.1) 7 (100)
122 (58.7) 86 (41.3)
0.0014 0.51 0.20 0.70
5 (71.4) 2 (28.6) 0.74
17 (8.2) 6 (2.9) 131 (63.0) 2 (1.0) 14 (6.7) 25 (12.0) 10 (4.8) 3 (1.4) 14 (6.7) 61 (29.3) 9 (4.3) 12 (6.1) 150 (134, 182) 78.5 (72, 101) 32 (15.4)
1 (14.3) 0 (0) 5 (71.4) 0 (0) 1 (14.3) 0 (0) 0 (0) 0 (0) 2 (28.6) 4 (57.1) 1 (14.3) 5 (71.4%) 351 (197, 391) 185 (93, 232) 5 (71.4)
a Continuous data are expressed as median (interquartile range) and categorical data as frequency (percentage). coronary artery patterns, see Table 1.
ASO ⫽ arterial switch operation;
p Value
0.09 0.20 0.29 0.0001 0.0003 0.0021 0.0019 b
For definitions of the various
TGA ⫽ transposition of the great arteries.
ing (LAD) and circumflex coronary artery, while the right coronary artery (RCA) was normal (variation 1a). Three other patients had close variations on this pattern. One patient had a single orifice (variation 1b), whereas another had separate origins from the rightfacing sinus and the circumflex arose from the RCA while the left anterior descending coronary artery ran an intramural course (variation 2). In the third patient, the origin of the intramural LCA was from above the posterior commissure (variation 3). The remaining 4 patients had different IMCA patterns. Two patients had a normally arising LCA but their RCA had a slitlike orifice at or above the sinotubular junction that traveled intramurally down the right-facing sinus before exiting normally onto the surface of the heart (variation 4). Two patients had bilateral IMCA patterns (variation 5). The first of these had bilateral origins above the sinotubular junction of the appropriate sinus that ran intramurally before exiting normally. The second of these had a single orifice well up the aorta to the right of the posterior commissure that immediately bifurcated to an intramural RCA that exited on the rightfacing sinus before giving off a circumflex branch, and an intramural left anterior descending coronary artery that took a long intramural course before exiting on the left-facing sinus.
Demographic Observations Of the 215 patients, 17 were referred from French Polynesia. Of these 17, 6 had IMCA, giving an overall prevalence of 35.3% in this population compared with 6.1% (12 of 198) in the remaining non–French Polynesian population (p ⫽ 0.001).
Early Outcome: Mortality There were 7 early deaths in our entire ASO cohort, giving an overall mortality of 3.26% (Table 3). There was 1 death in the IMCA population (5.56%) and 6 in the non-IMCA population (3.0%; p ⫽ 0.46). No single anatomic feature was associated with early mortality. However, when patients were grouped according to the presence or absence of concomitant cardiac comorbidity, statistical significance was reached. Concomitant cardiac comorbidity was defined as the presence of Taussig-Bing anomaly, or any aortic, pulmonary, or outflow tract abnormality requiring intervention, or low operative weight (1 kg or less). The additional complexities associated with these comorbidities are reflected in significantly longer cardiopulmonary bypass time, longer cross-clamp time, and a greater prevalence of unstable postoperative patients evident by delayed sternal closure. The older age at operation in the 7 who died reflects 3 premature neonates
PEDIATRIC CARDIAC
Characteristics
2088
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
Ann Thorac Surg 2012;94:2084 –90
Fig 3. Kaplan-Meier survival probability curves for intramural coronary artery pattern (IMCA) and non-IMCA cohorts.
PEDIATRIC CARDIAC
in whom treatment was delayed, and 2 others requiring transfer from Pacific Islands to New Zealand for treatment, both of whom died of sepsis attributable to Pseudomonas species. The neonate with the IMCA pattern who died had a Taussig-Bing type malformation. This neonate was transferred ventilated from Tahiti and underwent ASO at day 8 of life. Thirty-six hours postoperatively there was low cardiac output and leukopenia. The neonate was treated for sepsis but died several hours later. Post-mortem examination confirmed disseminated Pseudomonas infection.
Early Outcome: Morbidity (Intramural Coronary Artery Pattern) Intraoperatively, 1 of the 18 IMCA patients had ischemic changes on ECG when weaning from bypass. This patient was the first in our series and underwent repair using the pericardial hood technique. The ischemia was attributable to dynamic compression of the pericardial hood by the neopulmonary artery. This was resolved by elongation of the neopulmonary artery, plication of the posterior pericardial patch of the neoaorta, and the placement of anterior plication sutures on the aorta to change the angulation of the pericardial hood. Postoperatively, sternal closure was delayed in 5 patients. Two patients exhibited tachyarrhythmias that required chemical or electrical cardioversion. One patient experienced a focal seizure with no neurologic sequelae. Two patients had preoperative comorbidities that complicated their postoperative course. One of these had necrotizing enterocolitis. The other had an embolic cerebral infarct from femoral vein thrombosis after repeated femoral vein instrumentation before ASO. ELECTROCARDIOGRAM. One patient had ST-segment elevation
on ECG in the hours after surgery that subsequently resolved. There were no ischemic changes on ECG before discharge for all 17 surviving patients. All patients were in sinus rhythm with or without right bundle-branch block.
Before hospital discharge no patient had significant left or right ventricular dysfunction. One had mild neoaortic regurgitation; 4 had mild neopulmonary regurgitation. There were no residual atrial or ventricular septal defects found that required subsequent closure.
ECHOCARDIOGRAM.
Late Outcome: Mortality Follow-up for the entire cohort of 215 patients is 9.1 years (IQR, 6.0 to 11.4). There was no significant difference in survival. Intramural coronary artery pattern survival was 94.4% at 1, 5, and 10 years compared with 95.9%, 94.9%, and 94.3% during the same interval for the non-IMCA population (p ⫽ 0.99; Fig 3).
Late Outcome: Morbidity (Intramural Coronary Artery Pattern) After discharge each patient with IMCA was clinically assessed by a pediatrician or pediatric cardiologist and received ECG, echocardiogram, angiogram, and DSE. In some cases not all of these investigations were possible. Of the 17 surviving cases, 1 was lost to follow-up and only received an echocardiogram during an outpatient visit immediately after discharge. Of the remaining 16, 2 were from remote Pacific islands and did not return for catheterization or DSE as transfer was not considered justified in the context of being asymptomatic with normal ECG and echocardiogram. Another patient did not receive angiography or DSE because of unrelated illness. Two patients received angiography (which was normal) but did not receive DSE for reasons that are not documented. CLINICAL STATUS. Of the 16 patients who had long-term follow-up, all are asymptomatic in terms of their cardiovascular health (median time of follow up, 81.5 months; IQR, 50.5 to 103 months). ELECTROCARDIOGRAM. An ECG was available in 16 of 17 patients. There were no ischemic changes found at a
median period of 61 months (IQR, 37 to 77 months) after surgery. ECHOCARDIOGRAM. Sixteen patients received outpatient echocardiograms at a median period of 41 months (IQR, 23 to 77 months) after surgery. Left and right ventricular function was normal in all studies. Three patients exhibited neoaortic root dilation and mild or moderate aortic valve regurgitation, and 2 had mild neopulmonary regurgitation.
Nonselective aortic root angiography was undertaken in 13 patients at a median time of 13 months (IQR, 13 to 22 months) postoperatively. Coronary artery abnormalities warranting subsequent angiography were found in 2 patients. In 1 patient moderate left main stem stenosis was demonstrated at 22 months. A selective coronary angiogram at 91 months showed diffuse narrowing but no significant stenosis. The initial nonselective angiogram of the second patient occurred at 12 months postoperatively and showed mild stenosis of the LCA. Selective coronary angiogram at 42 months demonstrated a widely patent left coronary ostium.
ANGIOGRAPHY.
Eleven patients received DSE at a median time of 13 months (IQR, 12.5 to 28 months). All studies were negative for inducible ischemia. Two patients who did not receive DSE had normal angiography.
DOBUTAMINE STRESS ECHOCARDIOGRAM.
CORONARY INTERVENTIONS IN THE OVERALL ARTERIAL SWITCH OPERATION POPULATION. There have been no coronary interventions in the IMCA cohort to date and one intervention in the non-IMCA cohort (p ⫽ 1.0). This was in a child with the most common pattern who underwent ASO in 1996, and who had an internal mammary artery graft 8.5 years later.
Comment The prevalence of IMCA in this study is higher than previously noted in other reports [6, 11–14], including from our own institution [15]. In particular we have identified a subset of patients from French Polynesia with a high prevalence. The difference in prevalence between our current and previous reports is owing to the previous study describing an earlier period of time from 1984 to 1999 during which successful transfers from French Polynesia were less frequent. In addition, the current report focuses solely on a neonatal population. The reasons for the increased prevalence of IMCA in the French Polynesian cohort are not clear and are beyond the scope of the current paper; however, the finding raises the possibility of a genetic basis or alternatively a common environmental exposure. The predominant pattern we noted of an intramural LCA originating on the right-facing sinus and passing behind the posterior aortic commissure is consistent with
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
2089
that reported in a previous large series [6] and composed a significant number of the intramural cases to which Asou and colleagues originally applied the unroofing technique [11]. The presence of an intramural coronary artery did not carry an additional risk of early death in our population as has been found in other studies [5, 6, 15]. In fact, no single anatomic feature was found to be significantly associated with early mortality. However, additional cardiac comorbidities were significantly associated with early death. The additional operative burden associated with these comorbidities was reflected in significantly prolonged bypass and aortic cross-clamp times, and is an association seen in other studies [4, 15, 16]. The older age at operation seen in those who died is attributable to the prevalence of premature neonates in this group who had prolonged medical treatment before surgery, and to 2 neonates who were transferred from remote Pacific islands. Others have found that intramural coronary arteries are associated with coronary stenosis and coronary event after ASO [6, 8 –10]. For this reason we followed our patients with nonselective angiography and DSE routinely at 1 year after ASO. Our results showed few cases of coronary artery stenosis using the unroofing technique. Although we do not routinely follow up our general cohort with the same degree of structural evaluation unless clinically indicated, we note comparable and low rates of coronary reintervention procedures in both groups. Where stenosis was suspected in the IMCA group at 12 months by nonselective angiography, subsequent selective angiography demonstrated reduction in stenosis with time. This finding could be consistent with a remodeling process, or could reflect the inaccuracy of nonselective coronary imaging in infants. Given that angiography itself is not without risk, we now consider that 12-month follow-up angiography be deferred in patients with a negative DSE until an age when selective coronary angiography or high-resolution computed tomography is possible. In view of recent results with high-resolution computed tomography our current approach is to image noninvasively and then to follow with conventional selective coronary angiography if clinically warranted [17, 18].
Study Limitations Given our sample size it would take 3 deaths in the IMCA cohort (giving a mortality of 16.7%) compared with a mortality of 3% in the non-IMCA cohort for significance to be reached. The lack of significant difference in mortality we found may therefore be related to limited statistical power given the sample size.
Conclusions We report outcomes for a cohort of patients with IMCA who underwent ASO and who predominantly received coronary unroofing repair. There was a high prevalence of IMCA in this population (8.4%), especially in a subset of patients referred from Tahiti (35.3%). There was no
PEDIATRIC CARDIAC
Ann Thorac Surg 2012;94:2084 –90
2090
THRUPP ET AL ASO INTRAMURAL CORONARY OUTCOMES
significant difference in mortality between IMCA and non-IMCA patients. Patients are asymptomatic at a median follow-up time of 81.5 months with excellent structural and functional outcome as assessed by angiography and DSE. We have obtained acceptable early and late outcomes for patients with d-transposition of the great arteries or double-outlet right ventricle and an IMCA using the coronary unroofing technique.
PEDIATRIC CARDIAC
The authors gratefully acknowledge the contribution of Steve Withy, Mildred Lee and Felicity Roberts for their respective assistance with data acquisition, statistical analysis and manuscript preparation.
References 1. Freed DH, Robertson CMT, Sauve RS, et al. Intermediateterm outcomes of the arterial switch operation for transposition of great arteries in neonates: alive but well? J Thorac Cardiovasc Surg 2006;132:845–52. 2. Sarris GE, Chatzis AC, Giannopoulos NM, et al. The arterial switch operation in Europe for transposition of the great arteries: a multi-institutional study from the European Congenital Heart Surgeons Association. J Thorac Cardiovasc Surg 2006;132:633–9. 3. Armishaw J, Gentles TL, Calder AL, Raudkivi PJ, Kerr AR. Transposition of the great arteries: operative outcome in the current era. N Z Med J 2000;113:456 –9. 4. Kirklin JW, Blackstone EH, Tchervenkov CI, Castaneda AR. Clinical outcomes after the arterial switch operation for transposition. Patient, support, procedural and institutional risk factors. Congenital Heart Surgeons Society. Circulation 1992;86:1501–15. 5. Pasquali SK, Hasselblad V, Ji JS, Kong DF, Sanders SP. Coronary artery pattern and outcome of arterial switch operation for transposition of the great arteries: a metaanalysis. Circulation 2002;106:2575– 80. 6. Metton O, Calvaruso D, Gaudin R, et al. Intramural coronary arteries and outcome of neonatal arterial switch operation. Eur J Cardiothorac Surg 2010;37:1246 –53.
Ann Thorac Surg 2012;94:2084 –90
7. Legendre A, Losay J, Touchot-Koné A, et al. Coronary events after arterial switch operation for transposition of the great arteries. Circulation 2003;108(Suppl 1):II-186 –90. 8. Prêtre R, Tamisier D, Bonhoeffer P, et al. Results of the arterial switch operation in neonates with transposed great arteries. Lancet 2001;357:1826 –30. 9. Tamisier D, Ouaknine R, Pouard P, et al. Neonatal arterial switch operation: coronary artery patterns and coronary events. Eur J Cardiothorac Surg 1997;11:810 –7. 10. Bonhoeffer P, Bonnet D, Piéchaud J-F, et al. Coronary artery obstruction after the arterial switch operation for transposition of the great arteries in newborns. J Am Coll Cardiol 1997;29:202– 6. 11. Asou T, Karl TR, Pawade A, Mee RB. Arterial switch: translocation of the intramural coronary artery. Ann Thorac Surg 1994;57:461–5. 12. Hutter PA, Bennink GBWE, Ay L, Raes IB, Hitchcock JF, Meijboom EJ. Influence of coronary anatomy and reimplantation on the long-term outcome of the arterial switch. Eur J Cardiothorac Surg 2000;18:207–13. 13. Sachweh JS, Tiete AR, Jockenhoevel S, et al. Fate of intramural coronary arteries after arterial switch operation. Thorac Cardiovasc Surg 2002;50:40 – 4. 14. Qamar ZA, Goldberg CS, Devaney EJ, Bove EL, Ohye RG. Current risk factors and outcomes for the arterial switch operation. Ann Thorac Surg 2007;84:871–9. 15. Wong SH, Finucane K, Kerr AR, O’Donnell C, West T, Gentles TL. Cardiac outcome up to 15 years after the arterial switch operation. Heart Lung Circ 2008;17:48 –53. 16. Blume ED, Altmann K, Mayer JE, Colan SD, Gauvreau K, Geva T. Evolution of risk factors influencing early mortality of the arterial switch operation. J Am Coll Cardiol 1999;33: 1702–9. 17. Ou P, Mousseaux E, Azarine A, et al. Detection of coronary complications after the arterial switch operation for transposition of the great arteries: first experience with multislice computed tomography in children. J Thorac Cardiovasc Surg 2006;131:639 – 43. 18. Ou P, Celermajer DS, Marini D, et al. Safety and accuracy of 64-slice computed tomography coronary angiography in children after the arterial switch operation for transposition of the great arteries. JACC Cardiovasc Imaging 2008;1:331–9.