|
|
||||||||
Ann Thorac Surg 2005;79:1433-1444
© 2005 The Society of Thoracic Surgeons
a Division of Cardiovascular Surgery, Center For Congenital Heart Diseases, Children's Hospital, University of Zürich, Zürich, Switzerland
b Division of Cardiology, Center for Congenital Heart Diseases, Children's Hospital, University of Zürich, Zürich, Switzerland
* Address reprint requests to Dr Dodge-Khatami, Division of Cardiovascular Surgery, Children's Hospital, University of Zürich, Steinwiesstrasse 75, CH-8032 Zürich, Switzerland
ali.dodge-khatami{at}kispi.unizh.ch
| Abstract |
|---|
|
|
|---|
Abbreviations: ASO = arterial switch operation CCTGA = congenitally corrected transposition of the great arteries CHSS = Congenital Heart Surgeons Society ECMO = extracorporeal membrane oxygenation LV = left ventricle or left ventricular LVOTO = left ventricular outflow obstruction MRI = magnetic resonance imagery NR = not reported NYHA = New York Heart Association PA = pulmonary artery PHN = pulmonary hypertension RV = right ventricle or right ventricular SVC = superior vena cava TGA = transposition of the great arteries TI = tricuspid insufficiency VSD = ventricular septal defect y = years
| Introduction |
|---|
|
|
|---|
Without this operation, the natural history of patients with all variants of TGA was dismal, with 55%, 85%, and 90% mortality rates at 1 month, 6 months, and 1 year, respectively [2]. This ingenious procedure, also known as the atrial or venous switch, involves rerouting the pulmonary veins through the tricuspid valve to the systemic right ventricle (RV) by means of an atrial flap (fashioned from the free wall of the right atrium) plus the redirection of systemic venous blood from both vena cavae, through the mitral valve to the pulmonary left ventricle by using the intraatrial septum.
The initial results of this procedure were disappointing [3], as may be seen by the 7 hospital deaths from a series of 11 patients (63.7% mortality) that was reported by Kirklin and colleagues [4] in 1961. The high mortality and difficulty in reproducing Senning's own better experience [5] motivated others to modify the procedure. This ultimately lead to the Mustard operation in 1964 [6], in which a pericardial baffle was inserted. Quaegebeur and colleagues revived the Senning operation through technical modifications, resulting in considerable improvement of in-hospital survival [7].
Until the late 1970s, the atrial baffle operations were the only established procedures for the repair of complete transposition, and with increased experience, surgical mortality steadily decreased to low levels (1% to 9%) [8]. However, intermediate- to long-term survivors were being recognized with RV failure, systemic and pulmonary venous pathway leaks and obstructions, varying degrees of tricuspid valve insufficiency, atrial arrhythmias, and unexpected late sudden deaths.
Jatene successfully performed the first arterial switch operation (ASO) in 1975, which increasingly gained popularity, was reproducible with an acceptable learning curve, and resulted in lower mortality rates than the Senning operation. More important, it represented an anatomic and physiologic repair of transposition, placing the left ventricle (LV) in the systemic position, thus avoiding potential long-term RV failure that complicated the atrial baffle operations. This fact, and mortality that reaches zero in many centers that perform the ASO [9], have made the Senning operation a palliative procedure. As a result, the Senning operation has become nearly obsolete in the surgical management of neonates with TGA. It is important to note that the ASO transfers the pulmonary valve to the systemic position, with the potential for late neo-aortic valve incompetence, the long-term significance of which is still unknown.
Renewed interest in the Senning operation has emerged since the 1990s, as it is an essential part of the anatomic repair in patients with congenitally corrected transposition of the great arteries (CCTGA). Although this relatively new strategy achieves anatomic and physiologic repair of CCTGA, the number of large series is limited and the follow-up is short. Theoretically, the long-term complications that have been witnessed after a Senning operation for TGA could be anticipated after anatomic repair of CCTGA [10].
In this review we present technical details of the Senning operation, summarize the results of larger recent series, and address its potential long-term iatrogenic and physiologic implications. In a failing post-Senning heart, the indications for taking down an atrial baffle and retraining the left ventricle (LV) en route to an ASO will be discussed. It is hoped that the lessons learned from the atrial repair of TGA will serve us to better treat those patients with CCTGA who need an anatomic repair, in which the Senning operation has its newfound role.
| Material and Methods |
|---|
|
|
|---|
The data from our series and the ones presented in Table 1 and Table 2 were gathered from a retrospective chart review of all consecutive patients who had a Senning operation in each institution. Follow-up was performed by the respective cardiology teams in each hospital, by questionnaires sent to the patients residing outside of the country where the operation took place, or both. The term operative mortality in Table 1 refers to any death that occurred within 30 days of surgery.
|
|
| Results |
|---|
|
|
|---|
|
|
|
|
|
Long-Term Follow-Up Issues
EXERCISE TOLERANCE
The adequacy of the right ventricle to sustain the systemic circulation in the long-term can be questioned by its relative inefficient response to stress and effort, as illustrated by multiple studies enrolling patients after a successful atrial switch who are otherwise asymptomatic, in sinus rhythm, and without medication. Douard and colleagues performed bicycle ergometry in 43 asymptomatic patients at a mean follow-up of 11 ± 2.8 years after a Senning operation. They found reduced aerobic capacity, shorter exercise times, and lower maximal heart rates, indicating an impaired chronotropic response to effort [20]. Exercise capacity was inversely correlated with the time interval elapsed since surgery, suggesting that better functional results can be anticipated when the Senning operation is performed early. They also found an excessive ventilatory adaptation to exercise, reflected by an increased respiratory rate, a relative lesser increase in tidal volumes, and increased total ventilation, as compared to controls [20].
Matthys and colleagues pinpointed the lack of increase in stroke volume to be the underlying mechanism of an inefficient response to effort, stressing that RV dysfunction can exist without chronotropic impairment [21]. Also using bicycle ergometry, Gilljam and colleagues [22] demonstrated low oxygen uptake, low maximal heart rate, abnormal stroke volume response, and high total peripheral resistance in 17 adolescent patients after an atrial switch. The authors suggest contributing factors to include small and noncompliant atria with subsequent inadequate filling of the ventricles, ventilation-perfusion inequality, intrapulmonary shunts, and oxygen diffusion limitation between the alveoli and pulmonary capillaries [22].
Buheitel and colleagues [23] compared exercise performance of patients after a Senning operation or a Fontan completion with normal controls. They measured peak consumption of oxygen, maximal work rate, peak oxygen pulse, and end-expiratory pressure of carbon dioxide and found the poorest results in Fontan patients. The reaction to exercise was qualitatively identical between Fontan patients and those after a Senning operation, and comparable to that of patients with chronic heart failure. Quantitatively, they found the results of Senning patients to lie between controls and Fontan patients [23].
RIGHT VENTRICULAR FAILURE
After the atrial switch, the RV remains in the systemic circulation, similar to unoperated patients with CCTGA. Numerous reports have demonstrated the inadequacy of this ventricle to sustain the systemic circulation in the intermediate and long term, with RV dysfunction rates ranging between 4% and 16% [2429]. RV failure seems to be more prominent and occurs earlier in patients with TGA plus VSD, than in those with an intact interventricular septum [25, 27, 29]. RV failure is not a time-related event and can occur insidiously after a long period of apparent normal function in an otherwise asymptomatic patient [26]. This has been the major impetus towards not only abandoning the Senning operation for TGA but also for converting an atrial switch into an ASO and for promoting the anatomic repair in CCTGA, thus restoring the morphologic LV to the systemic circulation.
Using radionuclide ventriculography in 99 patients at a median of 13 years after an atrial switch, Reich and colleagues [27] demonstrated systolic dysfunction not only of the RV in 8% of patients, but also of the LV in 10% of patients. Diastolic dysfunction of the LV was present in up to 80% of patients and deteriorated with time [27]. Lubiszewska and colleagues [24] used myocardial perfusion imaging and radionuclide angiography to study 61 patients at rest and at exercise at a mean of 10 years after an atrial switch. Despite excellent exercise tolerance, RV systolic dysfunction was illustrated by a significantly reduced RV ejection fraction in all patients, mild perfusion defects in 14.7% of patients, and extensive perfusion abnormalities in 54% of patients, more often in the inferior and anterior wall of the RV. Perfusion abnormalities were more pronounced in patients who were older at the time of surgery and who had longer follow-up times. Also, moderate-to-severe tricuspid valve insufficiency was more frequent in patients with abnormal perfusion [24].
Confirming these results with a longer follow-up time of between 10 and 20 years after an atrial switch operation, Millane and colleagues [30] found perfusion defects in 21 of 22 patients studied (95%) at rest, during dipyridamole stress testing, or both. More alarming, these perfusion defects were irreversible in 55% of patients, indicating infarction or fibrosis, more importantly so in the anterior, inferior, and septal segments of the systemic RV. Concomitant wall-thickening abnormalities were noted in 83% of segments with fixed perfusion defects, mirrored by reduced wall motion [30].
Labbe and colleagues reported similar results in 43 patients 11.3 ± 3 years after a Senning operation by using thallium myocardial scintigraphy [31]. In a study comparing patients undergoing either a Senning operation or an ASO, Okuda and colleagues found reduced systolic shortening of the anteroposterior diameter of the systemic RV only in the Senning patients [32].
In unoperated patients with CCTGA, a morphologic RV sustains the systemic circulation and presents the same shortcomings as after a Senning correction. Hornung and colleagues [33] demonstrated reversible and fixed perfusion defects in 5 unoperated patients with CCTGA, correlating with regional wall motion, thickening abnormalities, and impaired RV contractility. Tulevski and colleagues found similar results in 13 adult patients with unoperated or physiologically repaired CCTGA by using magnetic resonance imagery (MRI) and dobutamine stress testing [34]. Both groups of authors conclude that ischemia and infarction are important causes of RV failure in patients with CCTGA, drawing parallels with the systemic RV after the atrial switch operation.
Somewhat contrary to this evidence, Lorenz and colleagues [35], using cine MRI, found markedly elevated RV mass, normal RV size, and only mildly depressed RV ejection fraction in 22 patients 8 to 23 years after an atrial switch procedure. Only 1 patient had clinical RV dysfunction with increased RV mass, a finding also observed in only 1 out of 40 patients in the series from Milwaukee [18]. They conclude that inadequate hypertrophy of the RV is not the cause of late RV dysfunction in patients after an atrial switch [35].
Using radionuclide cineangiography, Hochreiter and colleagues [36] studied 22 patients 8 to 18 years after an atrial switch and found not only normal resting RV and LV ejection fractions, but also preserved exercise endurance with normal RV ejection fraction at stress in patients having undergone their repair before the age of 1 year. They and others [18, 37] suggest that deleterious factors such as chronic hypoxia may explain the suboptimal results observed in older patients who undergo the atrial baffle procedure [36].
The cause of impaired RV function is presently unclear, and the available data are still inconclusive as to its implication. The etiology is probably multifactorial, either related to a late operation after chronic preoperative cyanosis and resultant RV ischemia, to suboptimal intraoperative myocardial protection, as was certainly the case in older series that used more primitive cardioprotective techniques, or to the inherent suboptimal geometry of the RV [28, 38]. Given the existence of adult patients whose RV volumes, function, and response to exercise are normal long after an atrial baffle procedure, it seems unreasonable to condemn the Senning or Mustard operations on the basis of inevitable RV dysfunction alone.
BAFFLE STENOSIS OR LEAK
Systemic vena cava stenosis corresponds to a pullback pressure difference of more than 5 mm Hg during catheterization [39]. Surprisingly, symptomatic caval obstruction is relatively rare, generally observed within weeks to several months after an atrial switch when it does occur, and rarely beyond 1 year postoperatively [39]. It is observed more frequently in patients who were operated on as neonates [16, 40, 41].
Superior caval obstruction is much more frequent than the obstruction of the inferior vena cava. When present, symptoms include puffiness of the eyelids or facial edema, pleural effusion, and even chylothorax [12]. Systemic venous obstruction has been reported more frequently after the Mustard operation (10% to 40% ) [15, 19] than after the Senning operation (0% )[25].
Pulmonary venous obstruction, contrary to systemic stenosis, is usually symptomatic. The reported incidence of this complication is 0% to 27%, much less frequently after the Senning operation [19, 39], although others have not found a statistical difference between the two procedures (10% after Mustard vs 13% after Senning) [40]. Symptoms consisting of cough, wheezing, dyspnea, and exercise intolerance usually present during the first year and indicate surgical reintervention, not infrequently on an urgent basis [19].
Baffle leaks (Fig 6) lead to residual interatrial shunts, either bidirectionnal or predominantly right to left. They are usually without hemodynamic significance and rarely indicate surgical reintervention for this reason alone [19]. Right-to-left shunting occurs in the absence of elevated systemic venous pressures and has to do with the streaming of blood underneath the interatrial baffle. The incidence ranges from 20% to 73% after the Mustard operation and from 0% to 50% after the Senning operation [39].
|
Deanfield and colleagues reported normal sinus rhythm in 84% of their patients in the immediate postoperative phase of an atrial baffle procedure, falling to 56% in stable sinus rhythm after a Senning correction, and to 66% after a Mustard operation, at a mean follow-up of 7 years [43]. They found no relation between the loss of sinus rhythm or active arrhythmia and sudden death, which occurs in up to 11% of patients as documented by Holter recordings [43]. In a more recent study from the same institution comparing the Senning and Mustard operations, the incidence of postoperative atrial flutter was similar and was strongly associated with late sudden death [12].
Intraatrial reentry tachycardia occurs in 2% to 10% of patients after the atrial switch operation [44]. It induces a rapid ventricular response and is thought to be one explanation for the 3% to 15% incidence of postoperative sudden death. Atrial tachyarrhythmias are induced by reentrant circuits that result from the extensive atrial suture lines involved in a Senning or Mustard operation [45]. Concealed entrainment techniques can be used to map reentry sites, which are most often found in the mouth of the coronary sinus and the tricuspid valve annulus, and in the atrial myocardium of right atrial origin, whether they are part of the surgically created pulmonary or venous atrium [45]. These sites can be successfully silenced with radiofrequency catheter ablation, and recurrence at midterm follow-up is low. This treatment modality aims to eliminate the electrical substrate for the arrhythmia and is hence more attractive than medication, which can result in breakthrough tachycardia or proarrhythmia, or both. Antitachycardia pacing has been used to treat intraatrial reentry tachycardia, but it carries the risk of accelerating the tachycardia into atrial fibrillation [45].
TRICUSPID VALVE INSUFFICIENCY
Various degrees of tricuspid valve insufficiency (TI) have been reported after the atrial switch, with an incidence that reaches as high as 52% in some series [8]. Relevant TI occurs more frequently after the Mustard correction than after a Senning operation [12]. It is more frequent in patients with TGA plus VSD, and may be related to intrinsic abnormalities of the tricuspid valve in these patients [25, 29, 46] or to intraoperative injury or distortion of the valve during VSD closure [25, 29, 39]. The incidence varies from 5% when the interventricular septum is intact to 30% with an associated VSD [19].
The degree of severity is usually mild, and symptoms or hemodynamic relevance are rare when TI occurs in the absence of RV failure [14,15, 39]. Accordingly, few reoperations are needed for isolated TI (see Table 2). According to Poirier and Mee [47], differences in outcome and eventual failure of the RV after a Senning procedure are related to the degree of TI in the immediate postoperative period, particularly in patients with TGA plus VSD. In the series from Melbourne [25], tricuspid valve damage at VSD closure or by jet lesions contributed to the difference in outcomes, suggesting that mild postoperative (post-Senning) TI could become significant and potentially lethal because it adds to the RV workload, further precipitating RV failure [25, 47].
When severe TI occurs, it is a precursor and near surrogate of impending RV failure, which it precedes by years [8, 29]. This may be addressed by tricuspid valve repair or replacement, although the results are disappointing [26, 28, 29, 37, 48], with minimum improvement in hemodynamics. When TI is associated with RV failure, atrial switch takedown, pulmonary artery (PA) banding and conversion to an ASO [28, 29], or transplantation, may be better options [8, 28, 48].
PULMONARY VASCULAR OBSTRUCTIVE DISEASE
The appearance and progression of pulmonary vascular obstructive disease and resultant pulmonary hypertension (PHN) in unoperated patients with TGA is related to age; the degree to which it persists, stabilizes, or regresses after an atrial switch is also a function of age at correction [39]. Mild-to-severe PHN occurs in 4% to 35% of patients with TGA plus an intact ventricular septum after an ASO [39]. This incidence is only 1% to 3% when the repair is performed before 1 year of age, and increases steadily thereafter. As a corollary, risk factors to develop PHN include older age at repair, the preoperative presence of a large patent ductus arteriosus, and a large VSD [39]. The progression of PHD after an atrial baffle procedure is rare but has been reported, as well as the even more rare decrease in pulmonary arteriolar resistance after surgical correction [39].
Psychosocial Outcomes and Neurodevelopment
After the various surgical repairs for TGA, children have more neurologic impairment, learning disabilities, behavioral disorders, and poorer motor and vocabulary abilities than their healthy peers [4951]. Although this has been extensively documented after the ASO by the group from Boston Children's Hospital [4951], the deficits found in this cohort do not seem specific to children with TGA, but are similar to those found in others undergoing repair of a congenital or acquired heart defect [51].
Alden and colleagues [52] studied 31 children who were operated on in one institution at a mean of 11.5 years after a Senning or Mustard repair, with varying cardiac functional status at last follow-up. Nineteen percent had a psychiatric diagnosis, mostly of an internalizing nature that tended to be predicted by the severity of the cardiac condition. This is still considerably lower than what has been reported after cardiac surgery for other cyanotic cardiac conditions. These children had good psychosocial functioning, and only one in five had severe emotional or behavioral problems. Intelligence quotient scores were marginally lower than the general population, but only one child was mentally retarded (3%) [52].
Culbert and colleagues [53] compared patients having undergone an ASO operation, a Senning or Mustard operation, and a Rastelli operation with healthy age-matched children. After TGA repair, children and adolescents functioned well both physically, and psychosocially. The complete patient population scored higher on the Child Health Questionnaires than control norms in all categories except self-esteem. Patients achieved higher scores after an ASO [53] than both subsets of patients undergoing an atrial baffle procedure. Contrary to this study, Ellerbeck and colleagues found no difference in cognitive and motor development, neurologic impairment, learning disabilities, behavior disorders, or motor, vocabulary and acquired abilities, between children after an ASO and an atrial switch operation [54].
Given the vast list of pre-, intra-, and postoperative variables that may affect the mid- to long-term neurodevelopmental status of a patient, it is currently difficult to establish whether the underlying disease itself, the type of surgical correction, or the technical aspects of cardiopulmonary bypass are responsible for the adverse outcomes [51].
Senning Versus Mustard
After the Senning procedure was abandoned in the mid-1960s and early 1970s in favor of the Mustard operation, renewed interest in the Senning procedure was gained after the technical modifications introduced and promoted by Quaegebeur and colleagues [7].
The theoretical and practical relative advantages of the Senning operation include avoidance of foreign material, potential for growth of native tissues forming the neo-chambers, potential functional capacity with muscular contraction of the atrial chambers, and avoidance of akinetic patches that can scar, shrink, thicken and further obstruct atrial inflow, such as that seen with the Mustard operation [41]. In a population-based cohort study that looked at mortality 25 years after surgery for congenital heart diseases, Morris and Menasche found an improvement in survival with the Senning operation compared with the Mustard operation (late cardiac mortality 2% at 10 years, and 15% at 15 years, respectively) [55]. Arrhythmias were a major cause of morbidity and mortality in survivors of the Mustard operation, but not with the Senning operation, after which no arrhythmia-related deaths were noted [55].
The recent multicenter study from Belgium compared the long-term outcome in 339 patients up to 30 years after one of the two atrial switch procedures [56]. Both groups had a relatively high early mortality rate, but actuarial survival at 10, 20, and 30 years was satisfactory at 91.7%, 88.6%, and 79.3%, respectively. This was slightly better for the Senning group, although not significantly. At late follow-up, Senning patients had better functional status, participated more actively in sports, and had fewer baffle-related problems than did the Mustard group [56].
Sarkar and colleagues [12] compared their series of 141 patients who underwent a Senning operation with 249 patients who underwent the Mustard operation during the same time period. Survival was significantly better for the Senning group, reinterventions for baffle-related problems or left outflow tract obstruction were significantly lower, and pacemaker insertion was less frequent. The loss of stable sinus rhythm was comparable in the two groups and unrelated to death. The incidence of atrial flutter was similar in both groups and strongly associated with late sudden death. The authors concluded that the Senning operation had superior results, with good late functional status, and argued that elective atrial baffle takedown and conversion to an ASO cannot be justified in asymptomatic post-Senning patients [12].
During the same historical period in which patients were enrolled to undergo either of the atrial baffle procedures, Helbing and colleagues [14] compared 60 patients after a Mustard operation with 62 patients after a Senning procedure. At respective median follow-up times of 16 and 11 years postoperatively, there were no differences with regards to baffle-associated problems, RV failure, sudden death, or functional status between the Mustard and Senning patients. Satisfactory long-term survival was similar, and excluding pacemaker implantation, no reoperations were necessary in either subset of patients. The only significant risk factor for the occurrence of sinus node dysfunction was the Mustard operation [14].
A technical pitfall of the Mustard operation involves the difficulty in shaping an appropriate baffle without creating systemic or pulmonary venous obstruction, particularly in neonates. In a meta-analysis that reviewed multicentric postoperative angiographic data, Graham [39] found both caval obstruction and pulmonary venous stenosis to be more frequent after the Mustard operation than after a Senning operation. Risk factors to develop systemic venous obstruction included the use of a Dacron (DuPont, Wilmington, DE) baffle, operation in early infancy (
6 months), and the use of a "trouser-shaped" baffle instead of a "dumbbell-shaped" baffle, such as that originally described by Mustard [39].
Contrary to these reports, The Congenital Heart Surgeons Society [11] found better early and late survival after the Mustard operation than after a Senning operation in a prospective cohort of patients with TGA who were destined to have either an ASO, or one of the two atrial switch procedures. Twenty-one patients who were intended to have an ASO had a Senning operation instead owing to unfavorable conditions or anatomy that were discovered in the operating room. This cross-over with higher risk patients undergoing the Senning operation may have influenced the difference in early survival in favor of the Mustard operation, but does not explain the difference in late survival. In the atrial switch subgroup, risk factors for long-term pacemaker requirement included patients with TGA plus VSD undergoing a Senning operation, and previous surgical atrial septectomy [11]. Institutional preference or experience could partially explain the better early and late results with the Mustard variation, although this is purely speculative.
LV Retraining and Senning Takedown En Route to an ASO
When RV failure after a Senning correction reaches an advanced stage, treatment options are limited to tricuspid valve replacement, orthotopic cardiac transplantation, or atrial baffle takedown and conversion to an ASO. As the first two procedures have their own set of disappointing results and long-term complications [25, 28, 29, 37], more groups advocate restoring the morphologic LV to the systemic circulation [25, 29, 57]. Most often, this cannot be done in one step, as the LV has accustomed itself to the low pressures found in the pulmonary circulation. Before a Senning or Mustard takedown and a successful ASO are attempted, the LV must be retrained. Pulmonary artery (PA) banding is required to achieve adequate LV muscle mass, as was first described by Mee [29].
Currently, there are no clear indications or discriminating points to decide when a patient should no longer be treated medically for heart failure, whether transplantation is deemed a better option, or whether one should directly proceed to LV retraining. This controversial topic finds proponents and adversaries for each therapeutic arm and may be institutional-based; its answer is beyond the scope of this review. As medical treatment and transplantation are well described in the literature, LV retraining en route to an ASO is briefly reviewed here.
Foremost, contraindications to LV retraining include irreversible LV dysfunction, pulmonary valve abnormalities that render it unsuitable as a future neo-aortic valve, LV outflow tract obstruction that cannot be relieved, and uncontrolled arrhythmias [47]. The response to LV retraining is poorer in patients who are older than 15 years, although a successful Senning takedown and ASO were performed in a 28-year-old patient [47]. The degree of preexisting RV failure does influence the response to LV retraining, owing to the common interventricular septum that bulges towards the LV that induces LV outflow tract obstruction and eventual LV failure at lower than expected LV pressures [25, 47]. For these reasons, earlier PA banding is advocated, before decongestive therapy for RV failure becomes necessary [25, 28].
The aims of PA banding are to achieve a LV/RV pressure ratio of 0.7 or greater. One or more bandings may be required over a period of approximately 1 year to induce adequate LV hypertrophy, although this period is generally shorter in younger children [47]. The preparatory stage of retraining is better tolerated in patients after a previous atrial baffle procedure than in patients with an unoperated TGA who present late [48]. The former do not require systemic-to-pulmonary shunts in addition to a PA band to maintain adequate saturations, as they already have a physiologic circulation [48]. PA banding can induce neo-aortic valve insufficiency [28, 48, 57, 58], and the relative cumbersome need to perform multiple operations to tighten or loosen a band before adequate LV retraining is achieved may promote wider applications for the new adjustable and teleguidable FloWatch-R-PAB (EndoArt SA, Lausanne, Switzerland) band [59].
Before debanding, Senning takedown, and conversion to an ASO, transthoracic echocardiography, cardiac catheterization, and MRI are performed. These seek to confirm a LV that generates more than 80% of systemic blood pressures at rest, suprasystemic pressures with isoproterenol, or normal LV mass and wall thickness, indexed for weight and age [47, 58]. The size of the coronary arteries, and namely, that of the left coronary artery before PA debanding, may influence the success of a subsequent ASO with regards to the increase in coronary flow reserve that is required to adequately perfuse the future systemic LV [60].
In appropriately selected patients, the results of the LV retraining protocol after a failed Senning en route to an ASO are good to excellent in prepubescent patients [25, 28, 47, 58], but give unpredictable results in patients older than 15 or 16 years [47, 58]. LV retraining has failed when inadequate LV hypertrophy or LV dysfunction occurs, or if atrial arrhythmias progress [47]. When LV retraining is unsuccessful with uncontrollable ongoing RV dysfunction before an ASO, or LV deterioration after a secondary ASO, then early transplantation should be considered [28, 48, 57, 58]. The results of transplantation for a failed atrial switch have been satisfactory when performed in a timely fashion, although the long-term consequences that are general to all posttransplant patients, namely issues pertaining to a lifelong immunosuppressive regimen, are of concern [28, 48].
Rebirth of the Atrial Baffle Procedures for Patients With CCTGA
Although the Senning operation seems outdated and is used only in exceptional cases to treat patients with TGA, increasing interest and experience is being gained with this procedure in patients with congenitally corrected transposition as part of the double switch or Senning-Rastelli procedures. These operations reposition the morphologic LV in the systemic circulation, also referred to as the "anatomic repairs" of CCTGA, and are currently the treatment of choice in patients with this anomaly [28, 47, 6166]. Most authors recommend anatomic repair when tricuspid valve regurgitation or RV dysfunction are present [62, 64]. Others are more aggressive and recommend anatomic repair for all patients with an adequate or trainable LV, although until which age this is feasible or gives acceptable results is still controversial [62, 65, 66].
Proponents of the anatomic repair have demonstrated better results when the double switch is performed, with or without prior LV training, before the age of 15 to 16 years [47, 62]. Results have been less satisfactory in older patients, and in some instances, the LV is simply no longer trainable, leaving transplantation as the only salvage alternative. Some controversy concerning the double switch still revolves around asymptomatic patients, with or without associated intracardiac defects [67]. Indeed, drawing parallels between unoperated patients with CCTGA and patients after atrial correction for TGA, normal or near normal RV function in the long-term has been demonstrated in minimally symptomatic or asymptomatic adult patients with CCTGA [67].
When an anatomic repair of CCTGA is performed, the Senning operation is the preferred atrial baffle procedure for most [47, 61, 63], although in the presence of dextrocardia, the Mustard operation may be technically easier to perform [10, 61]. The timing of an anatomic repair is based on the size of the VSD. When the VSD is restrictive, LV pressures remain low (infra-systemic), resulting in an untrained LV, and the procedure should be performed before 1 month of age [61]. If it is performed later, preliminary PA banding may be required to redevelop the LV. With a large VSD that results in unrestricted pulmonary blood flow and systemic PA pressures, the repair should be performed by 6 months of age to prevent the development of pulmonary vascular disease [61].
The results of this complex procedure are good to excellent, with mortality rates ranging from 0% to 15% [6265]. Long-term follow-up of the anatomic repair for CCTGA is still required for patients with valved conduits who have undergone a Senning-Rastelli procedure and for the aortic valve and the morphologic LV in patients after the double switch [64]. Although the LV is restored to the systemic circulation, the long-term complications related to the atrial part of the Senning operation, namely the venous pathway problems and atrial arrhythmias, may still be expected [10].
Indications to Perform an Atrial Switch Operation for TGA
There are still instances where the Senning operation may be indicated for patients with TGA. These include complex coronary anatomy precluding an ASO, or late referral in patients with TGA plus VSD, which is very commonplace in developing countries. In this situation, PNH and a LV that is inadequate or untrainable may both contraindicate an ASO [16]. Even in older infants with an intact ventricular septum and low LV pressures, there still may be a place for the atrial baffle procedure [37].
A certain subset of patients may be more common than reported, mostly in developing countries with suboptimal medical control and access to diagnosis, namely those with TGA plus an intact ventricular septum, and severe PHN without a correctable cause. In the absence of overt left-to-right shunting, idiopathic PHN tends to last well beyond the neonatal period, if it regresses at all, and is more difficult to manage with medical therapy.
Successful surgical correction has been achieved in neonates with TGA plus an intact septum and PHN with an ASO [68], at the cost of a lengthy and stormy postoperative course that required inhaled nitric oxide or even extracorporeal membrane oxygenation (ECMO) [69, 70]. In these patients, Sharma and colleagues reported 75% mortality with an attempted ASO [68]. Despite what seemed to be a "prepared" LV preoperatively, RV failure in the face of systemic pulmonary artery pressures resulted in death. In addition, neopulmonary valve insufficiency is also a well-documented possibility after an ASO and will worsen with poor right-sided hemodynamics in the face of PHN. In 6 similar infants presenting consecutively, they opted for a Senning repair that resulted in early extubation and hospital discharge as well as 100% survival. Four of the patients had normal pulmonary artery pressures at 1 year postoperatively. A morphologic LV is better suited to face systemic pulmonary pressures in the setting of patients with PHN, and may give better chances for survival [69].
The group from Great Ormond Street, London, has recently presented their evolving practice to expand the indications for an ASO, either for late referral or diagnosis, prematurity, or intercurrent illness [70]. In these difficult patients, increased experience and the availability of postoperative ECMO has allowed post-ASO survival in selected patients up to 6 months of age [70].
| Conclusions |
|---|
|
|
|---|
Surgeons can perform the Senning operation low mortality and minimal morbidity by applying technical modifications and paying meticulous attention to large and unobstructed venous pathways. The results of this procedure may be compared with the newer ASO for TGA with regards to initial operative success, although long-term complications of the atrial baffle procedure currently speak in favor of the ASO. Eventual RV failure is not a time-related event [26], and still hampers the late follow-up of patients after the Senning operation. There is some evidence that the onset or degree of RV failure [18, 24, 3638] or exercise intolerance [20, 36] may be reduced when the Senning operation is performed earlier, particularly before the age of 1 year [36]. Currently, no diagnostic tool exists that allows for prediction of eventual RV failure in patients after an atrial correction for TGA. Long-term arrhythmias remain a problem after the atrial switch. In some instances this may be treated conservatively, although more invasive radiofrequency catheter ablation is required in others. The insidious nature of the various arrhythmias and their potential but unproven relation to sudden death emphasizes the need for closer arrhythmia follow-up.
It is noteworthy that reports from Europe and Australia that compare the Senning and the Mustard operations point to better immediate and long-term results with the Senning operation. However, the successive meta-analyses from the Congenital Heart Surgeons Society, which enrolled North American centers, report better objective outcomes with the Mustard operation. This may only reflect schools of training that have historically favored one operation over the other, leading to increased and improved experience with each respective surgical procedure. Despite the theoretical advantage of avoiding foreign material in the Senning operation, one should ultimately proceed with what works best for each institution.
The superior results of the double switch or Senning-Rastelli operation compared with the "classic" or "physiologic" repair make the former the preferred surgical treatment in patients with CCTGA. Theoretically, eventual RV failure or tricuspid valve insufficiency should be avoided, as the morphologic LV and mitral valve are restored to the systemic circulation [63]. As the follow-up of the more modern anatomic repair is still short, atrial arrhythmias and venous pathway obstructions or leaks may still be expected [10], although the management of these problems may relatively be straightforward and without heavy dire consequences.
In the current era, the Senning operation in patients with TGA is reserved for those with unfavorable coronary anatomy, for late referral, or for patients with TGA and pulmonary vascular obstructive disease, even when referred at an earlier age. These situations and the choice for a Senning operation may be particularly frequent and pertinent in developing countries without access to nitric oxide or ECMO. In patients with CCTGA, the Senning operation is an integral part of the double switch or Senning-Rastelli operation, whose long-term follow-up is still awaited.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. W.C. Scherptong, H. W. Vliegen, M. M. Winter, E. R. Holman, B. J.M. Mulder, E. E. van der Wall, and M. G. Hazekamp Tricuspid Valve Surgery in Adults With a Dysfunctional Systemic Right Ventricle: Repair or Replace? Circulation, March 24, 2009; 119(11): 1467 - 1472. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |