In introduction to this review of the development and results of the arterial switch operation for discordant ventriculoarterial connections, it seems necessary to attempt a definition, in surgical historical terms, of the past, present, and future. Although the future may simply be defined as everything that will happen after tomorrow, defining the past is more subtle, and its border with the present is a vague and wider time zone shaped by human memory. Few complex congenital heart lesions have seen such a spectacular change in outcome occur as for discordant ventriculoarterial connections, owing to the development of bold and ingenious surgical techniques, rendering what was universally a lethal condition into one with survivors, after surgical correction with the arterial switch operation, currently expected to lead near-normal lives. The anatomy and even physiology of discordant ventriculoarterial connections were brilliantly described by Baillie and CliftReference Baillie and Clift 1 in 1799. Since that time, the road to the current status required brave attempts and temporary success at palliation, then unsuccessful trials of anatomic correction and abandoning the arterial switch, physiologic correction with the atrial (venous) switch, culminating in what today is regarded as the anatomic and physiologic correction of the defect, the arterial switch operation. Although fascinating subjects in themselves, the various attempts and successes of palliation such as the Blalock–Hanlon operation or the Senning and Mustard atrial baffle procedures will not be examined in this review. We focus solely on the arterial switch operation, and for the sake of simplicity, the past includes the earliest unsuccessful attempts at the arterial switch dating back to 1954 by William MustardReference Mustard, Chute and Keith 2 until 1975. The present could be summarised as the period between the first successful arterial switch operation, performed in 1975 by Jatene et al,Reference Jatene, Fontes and Paulista 3 until now. Finally, the future is open-ended, and we shall cover the important residual lesions facing our patients and the new challenges facing the medical community treating them.
The past: Mustard, Bailey, Kay, and Idriss – the early pioneers
During the early 1950s, mechanical cardiopulmonary bypass was still in its developing stages, cross-circulation championed by C. Walton Lillehei was making its debut, and William T. Mustard was using explanted monkey lungs in his cardiopulmonary bypass circuit.Reference Fazel, Williams and Goldman 4 , Reference Evans 5 Using biological oxygenators in animal experiments since 1951, Mustard was emboldened to attempt new surgical techniques in humans, stating that “…surgery in the human being could be undertaken in hopeless cases…Transposition of the great vessels was the condition selected”.Reference Mustard, Chute and Keith 2 In 1954, he reported on seven patients with discordant ventriculoarterial connections operated at Toronto's Hospital for Sick Children, of whom three underwent the arterial switch operation. A 6-week-old baby died of aortic suture line bleeding, a 19-day-old baby died of left coronary artery occlusion, and another 19-day-old baby died from ventricular fibrillation. Mustard transferred the left coronary artery only and felt that leaving the right coronary artery arising from the pulmonary artery would be compatible with longevity: all three died in the operating room.Reference Mustard, Chute and Keith 2 He noted two important anatomic features, which would play a key role in future successes of the operation: the size discrepancy between the great vessels and the variation in coronary artery patterns.Reference Mustard, Chute and Keith 2
The same year, Charles Bailey used hypothermia to perform the “switchover” operation at Hahnemann Hospital in Philadelphia.Reference Bailey, Cookson, Downing and Neptune 6 In this procedure, performed in three patients with simple transposition, the coronary arteries were not transferred, and no patient survived.Reference Bailey, Cookson, Downing and Neptune 6 A year later in 1955, Earle Kay attempted a similar procedure in two patients – a 4-week-old baby and a 3-month-old baby – at St. Luke's Hospital in Cleveland,Reference Kay and Cross 7 whereby both great vessels were transected, and after disconnecting both pulmonary arteries the pulmonary trunk was sutured to the underside of the aortic arch. The transected pulmonary arteries were then reconnected on both sides of the original ascending aorta, directly on the right, and with an interposition graft for the left pulmonary artery. The coronaries were not transferred, and both patients died.Reference Kay and Cross 7
Following experimental surgery in 19 dogs undergoing the “switchover” procedure with cardiopulmonary bypass, Farouk Idriss attempted the arterial switch in 1961 at Children's Memorial Hospital in Chicago in two patients, one 7 years old and another 3 months old.Reference Idriss, Goldstein, Grana, French and Potts 8 On pre-operative cardiac catheterisations, the left ventricular pressure was only one-third of the systemic right ventricular pressure in the first patient and was nearly equal in the second. His procedure involved transecting an intact cuff of the aorta with both coronary arteries still attached, flipping it upside down and transposing it to the left ventricular outflow tract, incorporating it as a slice-in between the left ventricle and the aorta.Reference Idriss, Goldstein, Grana, French and Potts 8 Both patients died in the operating room, from left ventricular failure in the older patient and from excessive bleeding in the second patient. In his discussion, Idriss comments on the importance of patient selection, noting that the “first patient died because the left ventricle was unable to handle the sudden load imposed on it…[and]…in spite of two clinical failures, we think this operation is worthy of further trials”.Reference Idriss, Goldstein, Grana, French and Potts 8
These early attempts required tremendous courage and ingenuity, but were met with failure and 100% mortality owing to many factors, for example imperfect bypass technology and gross suture material, unrelated to the correct and logical concept of the arterial switch. It led to abandoning the arterial switch contemporary to the first successful atrial or venous switch operation performed by Åke Senning in 1958.Reference Senning 9 , Reference Senning 10 Although the venous switch provided only physiologic repair, leaving the right ventricle in the systemic circulation, it was safe, reproducible, and rapidly became the standard operation for discordant ventriculoarterial connections with excellent short- to mid-term results. In this setting, pursuing or justifying a different approach required tenacity, more courage, vision to predict systemic right ventricular failure after the venous switch, and the need to surmount the ethical considerations pertinent to abandoning an existing successful surgical strategy.Reference Bull, Yates, Sarkar, Deanfield and de Leval 11 After further experimental work in calves and dogs by Anagnostopoulos in Chicago,Reference Anagnostopoulos 12 it would take another decade and a half for Adib Jatene to perform the first successful arterial switch operation in 1975 in Sao Paolo, Brazil.Reference Mustard, Chute and Keith 2 However, after his initial spectacular breakthrough, he reported five deaths in the addendum to his 1976 paper,Reference Jatene, Fontes and Paulista 13 making many surgeons hesitant to undertake such a daunting procedure instead of the safer Senning operation. Some continued to perform the switch without coronary transfer, with surprising initial success in case reports.Reference Danielson, Tabry, Mair and Fulton 14 It would eventually take more than another decade for the arterial switch to become the gold standard operation for the treatment of discordant ventriculoarterial connections. Where to draw the line between the past and the present is probably debatable: does the present era of the arterial switch begin in 1975 after a first success or in the late 1980s when it was more widely performed? In any event, the surgical history of the arterial switch operation is a fascinating illustration of how a surgical operation established itself based on logic, trial and error, and a visionary shift in long-term expectations.
The present
Currently, for discordant ventriculoarterial connections with intact ventricular septum, the arterial switch is electively undertaken between the first day and the first 2 weeks of life. For discordant ventriculoarterial connections with a ventricular septal defect, the repair is usually recommended by 2 to 3 months. The majority of initial successful arterial switch procedures were undertaken in older patients with a ventricular septal defect or a large patent arterial duct, whereby the left ventricle was pre-operatively still submitted to systemic pressures and was therefore able to handle its systemic role after an arterial switch. One of the initial successful series was reported by Pacifico et al.Reference Pacifico, Stewart and Bargeron 15 However, the majority of babies with discordant ventriculoarterial connections do not have a ventricular septal defect, and very early on in the experience it was recognised that an unprepared left ventricle in patients with an intact ventricular septum would fail. The concept of two-stage anatomic correction was proposed by Yacoub et alReference Yacoub, Radley-Smith and Maclaurin 16 in 1977 and included pulmonary arterial banding followed by debanding and an arterial switch. This initial approach involved a period of banding of a year or more, further prolonging the duration of cyanosis, and imposing surgical morbidity such as time-related left ventricular injury, branch pulmonary arterial distortion, and insufficiency of the neoaortic valve.Reference Jonas, Giglia, Sanders, Wernovsky and Castaneda 17 , Reference Boutin, Jonas, Sanders, Wernovsky, Mone and Colan 18 The concept was later refined in an accelerated rapid, two-stage strategy by Jonas in 1989, by banding with or without an aortopulmonary arterial shunt, proving the feasibility of retraining the left ventricle by doubling its mass through hypertrophy after a week only.Reference Jonas, Giglia, Sanders, Wernovsky and Castaneda 17 , Reference Boutin, Jonas, Sanders, Wernovsky, Mone and Colan 18 The current guidelines indicating the adequacy of left ventricular training after pulmonary arterial banding include an inferior limit of left ventricular mass at 35 g/m2,Reference Lacour-Gayet, Piot and Zoghbi 19 a left-to-right ventricular pressure ratio >0.85, left ventricular end-diastolic volume >90% of normal, or posterior wall thickness >4 mm, allowing for a safe arterial switch after debanding.Reference Nakazawa, Oyama and Imai 20 In developing countries and/or in situations with late diagnosis and referral, the time limits to perform an arterial switch with or without prior left ventricular training by banding are being redefined and vary depending on the presence or absence of a ventricular septal defect,Reference Kang, de Leval and Elliott 21 – Reference Al Qethamy, Aizaz, Aboelnazar, Hijab and Al Faraidi 24 details of which are beyond the scope of this review. It is worth noting, however, that the age boundaries to directly perform an arterial switch without prior banding in babies with an intact ventricular septum have been pushed to 6 months.Reference Kang, de Leval and Elliott 21 This strategy needs to take into consideration the potential need for post-operative mechanical circulatory support during the recovery period.Reference Kang, de Leval and Elliott 21 In patients with a ventricular septal defect with or without Taussig–Bing morphology, patients of up to 19 years of age have effectively been treated by a direct arterial switch operation and closure of all defects, including a post-operative protocol of inhaled nitric oxide and oral sildenafil or bosentan.Reference Liu, Hu and Shen 23
Concerns regarding ongoing attrition of untreated patients with discordant ventriculoarterial connections and intact ventricular septum and/or the added surgical morbidity of the two-stage approach motivated Paul Ebert and the San Francisco group to embark on a brave approach to performing the neonatal arterial switch operation, amongst other neonatal repairs.Reference Turley, Mavroudis and Ebert 25 Early on, two of the first six babies died, one required reoperation for pulmonary artery stenosis, and another had neoaortic valve insufficiency, leading him to discontinue the programme temporarily. The LeCompte manoeuvre was introduced in 1982, making a case for reconstructing the right outflow tract without a prosthetic conduit.Reference Lecompte, Neveux and Leca 26 In 1984, using the LeCompte manoeuvre, Castaneda et al reported the first encouraging results in 14 neonates with transposition and intact septum with only one death, making a strong case for an early primary arterial switch while the left ventricle is still capable of effectively sustaining the systemic circulation.Reference Castaneda, Norwood, Jonas, Colon, Sanders and Lang 27 From this point in the mid-1980s, the stage of the present era was finally set: the indications and timing relating to anatomic subtypes were better defined, the surgical technique was refined, the coronary artery anatomic variations were better understood, cardiopulmonary bypass was up to par, and neonatal intensive care improved. Various international groups could define the standard of care and current expectations for a lesion, which only three decades prior had eluded all hope of treatment. In the present era, which has continued for 25 years, operative mortality for each anatomic subtype – simple transposition, complex transposition including a ventricular septal defect, arch obstruction, coarctation of the aorta, Taussig–Bing anomaly, and intramural coronaries – is clearly defined,Reference Mavroudis 28 – Reference Williams, McCrindle, Ashburn, Jonas, Mavroudis and Blackstone 40 and surgical performance can be measured. Hospital mortality for simple transposition is up to 5% and for complex transposition is 5–10%.Reference Mavroudis 28 – Reference Williams, McCrindle, Ashburn, Jonas, Mavroudis and Blackstone 40 With these figures as a benchmark, it is easy to see that the vast majority of babies with discordant ventriculoarterial connections are expected to survive the arterial switch operation with a heart that is cured. Or is it really? In 1990, Kirklin et alReference Kirklin, Colvin, McConnell and Bargeron 41 and othersReference Castaneda, Norwood, Jonas, Colon, Sanders and Lang 27 – Reference Yacoub 29 were able to foresee the possibility of morbidity or residual lesions to be found after the arterial switch operation, making it perhaps the best surgical solution currently available for patients with discordant ventriculoarterial connections but certainly not a cure.
An unusual but increasingly encountered indication for the arterial switch operation is for patients with discordant ventriculoarterial connections and a failing systemic right ventricle after a Senning or Mustard-type atrial switch.Reference Mavroudis and Backer 42 – Reference Poirier, Yu, Brizard and Mee 45 Indeed, the two most common and feared mid- to long-term complications after the atrial baffle procedure are intractable atrial arrhythmia and systemic right ventricular failure,Reference Mavroudis and Backer 42 – Reference Poirier, Yu, Brizard and Mee 45 estimated at ∼10% after the first decade and expected to increase with time.Reference Turina, Siebenmann, von Segeser, Schönbeck and Senning 43 , Reference Poirier, Yu, Brizard and Mee 45 In this setting, as with unoperated patients with discordant ventriculoarterial connections referred late for an arterial switch, a chronically underloaded left ventricle cannot be expected to suddenly handle a systemic workload if (re)-switched back into the systemic circulation. The concept of left ventricular retraining by pulmonary arterial banding before an atrial baffle takedown, debanding, and arterial switch operation was introduced by Roger B Mee in 1986.Reference Mee 44 To reach this goal, one or multiple pulmonary arterial banding operations and a period of 1–5 years is thought to be required to achieve adequate left ventricular retraining.Reference Mavroudis and Backer 42 , Reference Mee 44 , Reference Poirier, Yu, Brizard and Mee 45 Echocardiographic and cardiac catheterisation criteria to suggest adequate left ventricular function before an arterial switch include a left ventricular posterior wall thicker than 8 mm during diastole and a left-to-right ventricular systolic pressure ratio >80%.Reference Mavroudis and Backer 42 At the time of debanding and arterial switch, concomitant antiarrhythmia surgery and dealing with neoaortic valve insufficiency should be applied as appropriate, despite an already extensive surgical procedure.Reference Mavroudis and Backer 42 The staged strategy has met with success in a very select group of younger children with good long-term outcomes, and its indication may be broadening. However, complex coronary anatomy and older age beyond 12–16 years are reported risk factors for mortality or failure to complete the full protocol.Reference Mavroudis and Backer 42 , Reference Poirier, Yu, Brizard and Mee 45 In this case, the only alternatives are to leave the banding as destination therapy or proceed to cardiac transplantation.Reference Mavroudis and Backer 42 , Reference Poirier, Yu, Brizard and Mee 45 It is therefore crucial but very difficult to distinguish between adolescents with discordant ventriculoarterial connections who will continue to do well long after a Senning or Mustard palliation, and others to commit early enough to a left ventricular retraining protocol, in view of debanding and a successful arterial switch operation.Reference Mavroudis and Backer 42 , Reference Poirier, Yu, Brizard and Mee 45
Predictable residuals and potential problems for the future
There exist predictable but treatable surgical problems pertaining to the pulmonary arteries, the neoaortic valve, and the coronary arteries, as well as more obscure residua involving the myocardium, coronary reserve, exercise capacity, and neurodevelopmental issues faced by patients long after the arterial switch operation, with future true implications still unknown.
Pulmonary artery/right ventricular outflow tract stenosis
Right ventricular outflow tract obstruction, more specifically supravalvar pulmonary stenosis, is the most common mid- to long-term complication after the arterial switch operation, occurring in ∼2–17%Reference Fricke, d'Udekem and Richardson 32 – Reference Oda, Nakano, Sugiura, Fusazaki, Ishikawa and Kado 35 of cases. Despite recent advances in reconstruction technique and the more liberal use of fresh autologous pericardium to reconstruct the pulmonary arteries, it still represents about half of all indications for reintervention/reoperation after the arterial switch, independently of the material used for reconstruction.Reference Lecompte, Neveux and Leca 26 Discordant ventriculoarterial connections with a ventricular septal defect,Reference Fricke, d'Udekem and Richardson 32 , Reference Rudra, Mavroudis and Backer 34 prior neopulmonary artery reconstruction with a Gore-Tex patchReference Rudra, Mavroudis and Backer 34 or an equine pericardial patchReference Oda, Nakano, Sugiura, Fusazaki, Ishikawa and Kado 35 have been charged with increasing the risk for a reintervention on the right ventricular outflow tract. Fortunately, supravalvar pulmonary stenosis is easy to treat and can in the vast majority of cases be performed in the interventional catheterisation laboratory by percutaneous balloon dilatation.Reference Fricke, d'Udekem and Richardson 32 – Reference Oda, Nakano, Sugiura, Fusazaki, Ishikawa and Kado 35 As it is almost always amenable to curative dilatation or surgical correction in the rare cases requiring a reoperation, it is not expected to represent a major concern for ongoing or progressive morbidity in the long-term for these patients.
Neoaortic valve
After the arterial switch, the structurally formed pulmonary valve becomes the neoaortic valve. Very early on in the experience, concerns were raised as to the longevity of the neoaortic valve in its new role.Reference Castaneda, Norwood, Jonas, Colon, Sanders and Lang 27 , Reference Mavroudis 28 Although severe obstruction of the left ventricular outflow tract or organic narrowing of the pulmonary valve are still considered contraindications to perform the arterial switch operation,Reference Moulton, de Leval, Macartney, Taylor and Stark 46 there are no obvious criteria to proceed with or discard the arterial switch with borderline anatomy of the future neoaortic valve, including in patients with a bicuspid pulmonary valve, present in 1–7% of patients with transposition.Reference Uemura, Yagihara and Kawashima 47 – Reference Sohn, Brizard, Cochrane, Wilkinson, Mas and Karl 50 However, there seems to be increasing evidence that a bicuspid pulmonary valve is not a contraindication to perform an arterial switch, as demonstrated by mid- to long-term follow-up data revealing a low prevalence of neoaortic valve insufficiency or stenosis, comparable with that after switching a normal trileaflet pulmonary valve.Reference Uemura, Yagihara and Kawashima 47 – Reference Sohn, Brizard, Cochrane, Wilkinson, Mas and Karl 50 So how good is the durability of the neoaortic valve in general? Answering this question may be relative in terms of documenting any degree of valve stenosis/regurgitation or significant neoaortic valve dysfunction requiring a reoperation.Reference Fricke, d'Udekem and Richardson 32 – Reference Oda, Nakano, Sugiura, Fusazaki, Ishikawa and Kado 35 , Reference Williams, McCrindle, Ashburn, Jonas, Mavroudis and Blackstone 40 , Reference Angeli, Gerelli and Beyler 49 – Reference Pan, Hu and Li 62 In mid-term follow-up studies, the incidence of mild aortic regurgitation occurs in 13–38% of patients, moderate aortic regurgitation is reported in 0.7–15%, and severe aortic regurgitation in only 0.4–1.5%Reference Fricke, d'Udekem and Richardson 32 , Reference Lalezari, Bruggemans, Blom and Hazekamp 33 , Reference Vandekerckhove, Blom, Lalezari, Koolbergen, Rijlaarsdam and Hazekamp 52 , Reference Lange, Cleuziou and Hörer 61 of patients. Freedom from at least moderate aortic insufficiency is 100%, 96–100%, 94–99%, 91–99%, and 85–98% at 1, 5, 10, 15, and 20 years, respectively.Reference Fricke, d'Udekem and Richardson 32 , Reference Losay, Touchot and Serraf 54 , Reference Lange, Cleuziou and Hörer 61 Accordingly, the reoperation rate for aortic valve repair or replacement is very low, reported at 0.3–5%.Reference Fricke, d'Udekem and Richardson 32 , Reference Lalezari, Bruggemans, Blom and Hazekamp 33 – Reference Oda, Nakano, Sugiura, Fusazaki, Ishikawa and Kado 35 , Reference Williams, McCrindle, Ashburn, Jonas, Mavroudis and Blackstone 40 , Reference Angeli, Raisky, Bonnet, Sidi and Vouhé 51 , Reference Losay, Touchot and Serraf 54 , Reference Marino, Wernovsky and McElhinney 55 , Reference Choi, Kwon and Kim 60 , Reference Lange, Cleuziou and Hörer 61 Risk factors predicting neoaortic regurgitation include prior pulmonary artery banding,Reference Rudra, Mavroudis and Backer 34 the trap-door technique for coronary reimplantation,Reference Formigari, Toscano and Giardini 53 complex coronary anatomy,Reference Rudra, Mavroudis and Backer 34 an augmented sinotubular junction/aortic root dilatation,Reference Jhang, Shin and Park 56 , Reference Lim, Kim, Lee and Kim 57 , Reference Pan, Hu and Li 62 size discrepancy of the great arteries,Reference Lim, Kim, Lee and Kim 57 complex form of discordant ventriculoarterial connections – associated with ventricular septal defect, Taussig–Bing, or arch obstruction,Reference Rudra, Mavroudis and Backer 34 , Reference Oda, Nakano, Sugiura, Fusazaki, Ishikawa and Kado 35 , Reference Pan, Hu and Li 62 age greater than 6 months at the time of arterial switch surgery,Reference Pan, Hu and Li 62 and left ventricular outflow tract obstruction.Reference Lange, Cleuziou and Hörer 61 It is reassuring that despite a relative high incidence of neoaortic valve insufficiency, it is often mild, progresses but not severely, and only rarely leads to aortic valve reoperation.
Coronary arteries
Coronary stenosis after the arterial switch may be a life-threatening lesion, and its true incidence is unknown.Reference Ou, Khraiche and Celermajer 63 , Reference El-Segaier, Lundin, Hochbergs, Jögi and Pesonen 64 Most patients are asymptomatic, as they feel no chest pain owing to the denervation of the heart during the arterial switch operation.Reference El-Segaier, Lundin, Hochbergs, Jögi and Pesonen 64 Standard electrocardiograms or echocardiography usually show no sign of myocardial ischaemia.Reference Angeli, Raisky, Bonnet, Sidi and Vouhé 51 , Reference El-Segaier, Lundin, Hochbergs, Jögi and Pesonen 64 – Reference Pasquali, Marino, McBride, Wernovsky and Paridon 69 Furthermore, progressive stenosis of one ostium may allow extensive collateralisation from the other coronary artery, thereby hindering ischaemia or delaying it in an unpredictable manner.Reference Mavroudis, Stewart, Backer, Rudra, Vargo and Jacobs 68 Almost all authors agree that diagnostic imaging is immediately necessary in a symptomatic patient, when there is any doubt as to whether an electrocardiogram exists, or after difficult coronary transfer. However, in a completely asymptomatic patient with prior standard coronary anatomy and an uneventful post-operative course, no consensus exists as to when and how often screening should be performed with elective coronary catheterisation, multi-slice computed tomography, or magnetic resonance imaging angiography. Independent of the coronary anatomy, some centres recommend systematic coronary angiography for all patients in the first 2–3 years after repair,Reference Angeli, Formigari and Pace Napoleone 65 before the age of 4–5 years,Reference Losay, Touchot and Serraf 54 , Reference Raisky, Bergoend and Agnoletti 70 , Reference Agnoletti, Bajolle, Bonnet, Sidi and Vouhé 71 or at 5, 10, and 15 years of age.Reference El-Segaier, Lundin, Hochbergs, Jögi and Pesonen 64 , Reference Bonnet, Bonhoeffer and Piéchaud 66 Myocardial perfusion scanning has been recommended in all asymptomatic patients at 3 years.Reference Agnoletti, Bajolle, Bonnet, Sidi and Vouhé 71
The reported incidence of late coronary stenosis or occlusion ranges from 2% to 9%.Reference Angeli, Raisky, Bonnet, Sidi and Vouhé 51 , Reference Losay, Touchot and Serraf 54 , Reference Lim, Kim, Lee and Kim 57 , Reference Ou, Khraiche and Celermajer 63 – Reference Agnoletti, Bajolle, Bonnet, Sidi and Vouhé 71 Mechanisms leading to stenosis include ostial fibrosis at the suture line, mechanical kinking or stretching, or reactive injury to surgical manipulation.Reference Mavroudis, Stewart, Backer, Rudra, Vargo and Jacobs 68 Risk factors for coronary events or the need for reoperation include single coronary arteryReference Lim, Kim, Lee and Kim 57 and complex coronary anatomy.Reference Rudra, Mavroudis and Backer 34 , Reference Angeli, Formigari and Pace Napoleone 65 Operative techniques include coronary ostial patchplasty with autologous or biological patch material, and/or coronary artery bypass grafting.Reference Rudra, Mavroudis and Backer 34 , Reference Angeli, Formigari and Pace Napoleone 65 , Reference Raisky, Bergoend and Agnoletti 70 Interventional balloon angioplasty and stenting have also been performed with success.Reference El-Segaier, Lundin, Hochbergs, Jögi and Pesonen 64 , Reference Hausdorf, Kampmann and Schneider 72 As the oldest survivors of the arterial switch are reaching adulthood, an age where natural atherosclerotic disease may affect the coronaries, it remains unclear whether or not the superimposed surgical trauma in the newborn period accrues risk in the long term.Reference Mavroudis, Stewart, Backer, Rudra, Vargo and Jacobs 68 It would seem wise to recommend lifetime follow-up at regular intervals with precise coronary imaging, in order not to miss silent coronary lesions with life-threatening consequences. However, clear guidelines have not been clarified.
Myocardial function/exercise capacity
Even in the presence of documented patent coronary arteries after the arterial switch, it is important to remember that the myocardium may have suffered during the initial surgical procedure or in the immediate post-operative period after a difficult recovery. It is possible that various baseline coronary patterns may increase risk in the long term, despite adequate surgical transfer, leading to progressive coronary insufficiency. Abnormal myocardial perfusion scans at rest and at peak exercise have been demonstrated in otherwise healthy and asymptomatic patients long after the arterial switch,Reference Weindling, Wernovsky and Colan 73 associated with diminished exercise capacityReference Tobler, Williams and Jegatheeswaran 59 , Reference Sterrett, Schamberger, Ebenroth, Siddiqui and Hurwitz 74 and mildly diminished peak heart rates, oxygen pulse and aerobic capacity.Reference Sterrett, Schamberger, Ebenroth, Siddiqui and Hurwitz 74 Sterrett et alReference Sterrett, Schamberger, Ebenroth, Siddiqui and Hurwitz 74 found this to be related to variant coronary artery patterns, in contrast to normal exercise perfusion scans documented in patients with normal coronary patterns. In the absence of a surgically correctable coronary ostial stenosis, this type of coronary insufficiency during stress is still important to diagnose in the early teenage years,Reference Sterrett, Schamberger, Ebenroth, Siddiqui and Hurwitz 74 as patients may benefit from counselling to limit strenuous competitive sports. Confirming these findings is a study by Pasquali et al, using exercise stress testing in 53 patients at a median of 14.1 years (range, 7.7–20.6 years) after an arterial switch operation.Reference Pasquali, Marino, McBride, Wernovsky and Paridon 69 They found variant coronary patterns – circumflex from right coronary artery, inverted right coronary artery and circumflex, single and/or intramural coronary – and a ventricular septal defect to be associated with chronotropic impairment of the heart and a trend towards lower percent of predicted peak oxygen consumption.Reference Pasquali, Marino, McBride, Wernovsky and Paridon 69 In another study excluding the potential negative bias of overprotection by parents and child deconditioning from reduced daily-life activity, van Beek et alReference van Beek, Binkhorst and de Hoog 75 also found significantly reduced peak exercise capacity, diminished maximal workload, and slower maximal heart rates in a group of young teenagers after the arterial switch, compared with healthy controls. Using echocardiography, Hui et alReference Hui, Chau, Leung, Chiu and Cheung 76 studied 31 asymptomatic children ∼10 years of age after the arterial switch with patent coronary artery ostia and found lower fractional shortening, ejection fraction, and rate-corrected velocity of circumferential fibre shortening. Alarmingly, dobutamine stress testing further unmasked wall motion abnormalities in 75% of patients, corresponding to reversible myocardial perfusion defects as detected by exercise perfusion scanning.Reference Hui, Chau, Leung, Chiu and Cheung 76 In the absence of well-defined coronary artery territorial involvement, the authors suggest global impairment of left ventricular function from diffuse coronary insufficiency, which was associated with older age at operation, longer cardiopulmonary bypass and circulatory arrest times, and unusual coronary patterns.Reference Hui, Chau, Leung, Chiu and Cheung 76 The clinical relevance and eventual therapeutic consequences of these findings remain unclear, although all authors mentioned previously recommend further follow-up of left ventricular function with regular myocardial perfusion scanning.Reference Tobler, Williams and Jegatheeswaran 59 , Reference Pasquali, Marino, McBride, Wernovsky and Paridon 69 – Reference Hui, Chau, Leung, Chiu and Cheung 76
Quality of life/neurologic/developmental outcomes
Reports on the perceived quality of life have shown different results depending on the age group studied.Reference Müller, Hess, Hörer and Hager 77 – Reference de Koning, van Osch-Gevers and Ten Harkel 80 Although self-reported quality of life and the degree of social independence seems excellent in the majority of teenagers and young adults after the arterial switch operation,Reference Müller, Hess, Hörer and Hager 77 – Reference Culbert, Ashburn and Cullen-Dean 79 the same has not been found in younger patients.Reference de Koning, van Osch-Gevers and Ten Harkel 80 In their study of 49 11-year-old arterial switch survivors, de Koning et alReference de Koning, van Osch-Gevers and Ten Harkel 80 reported poorer health-related quality-of-life perception with regard to motor functioning and positive emotional functioning, compared with age-matched healthy peers. More specifically, neurodevelopmental assessment in teenagers 16 years after the arterial switch shows satisfactory average scores compared with healthy controls, although consistently below expected levels, independent of the bypass strategy chosen at the time of surgery.Reference Bellinger, Wypij and Rivkin 81 In all, one-third required tutoring; a quarter had received special education, occupational therapy, or psychotherapy; and one in six had repeated a grade level at least once.Reference Bellinger, Wypij and Rivkin 81 Normal coping mechanisms towards young adulthood include a change in self-perception, explaining the perceived normalisation of the quality of life from the teenage years into adulthood.Reference Müller, Hess, Hörer and Hager 77 However, teenagers after the arterial switch operation have been shown to lack insight into their weaknesses,Reference Bellinger, Wypij and Rivkin 81 and given their relative unawareness with regard to social cognition these adolescents should probably remain under surveillance, so as to detect and hopefully influence/coach their future chances of occupational success.Reference Bellinger, Wypij and Rivkin 81
Conclusion
The arterial switch operation remains the procedure of choice for patients with discordant ventriculoarterial connections when performed in the first days to weeks of life. Excellent early and long-term survival is now expected after successful surgery for a lesion that 50 years ago was considered fatal. Although the correction is deemed curative, certain residual anatomic lesions remain, leading to predictable reinterventions and reoperations, including relief of supravalvar pulmonary stenosis, left ventricular outflow tract reconstruction with aortic valve repair/replacement, and myocardial revascularisation. The overall reintervention rate is low, with freedom from a major event at 10 years nearing 90%.Reference Yamazaki, Yamamoto and Sakamoto 78 These are often difficult reoperations, requiring tremendous experience and surgical expertise, but results are excellent in capable hands. Other functional residua, such as global coronary insufficiency, diminished exercise capacity, and neurodevelopmental limitations, of which incidence and full clinical implications are still uncertain, need to be closely followed. After a more difficult initial adaptation period in children and adolescents,Reference de Koning, van Osch-Gevers and Ten Harkel 80 the quality of life is very satisfying in young adults.Reference Müller, Hess, Hörer and Hager 77 – Reference Culbert, Ashburn and Cullen-Dean 79
