| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1995;60:658-664
© 1995 The Society of Thoracic Surgeons
The Cardiothoracic Sciences Centre, All India Institute of Medical Sciences, New Delhi, India
Accepted for publication February 24, 1995.
Abstract
Background. Rapid two-stage arterial switch operation is advocated in infants with simple transposition presenting late. Accurate assessment of left ventricular preparation is crucial to successful outcome. The role of echocardiography alone in surgical decision making remains unclear.
Methods. Seventeen patients with simple transposition (mean age, 4 months) underwent pulmonary artery banding and modified Blalock-Taussig shunt (first stage) to prepare the left ventricle for the arterial switch operation (second stage). Serial echocardiography was performed in the interval phase to assess left ventricular growth. Sixteen patients underwent arterial switch operation after a mean interval of 10.4 ± 4 days, with 14 successful conversions. There was one mortality (5.9%) and two conversions to a Senning repair.
Results. In all patients a mean increase in left ventricular mass (40.8 ± 17.8 g/m2 to 81.4 ± 25.4 g/m2) and posterior wall thickness (3.37 ± 0.47 mm to 4.63 ± 0.58 mm) was recorded. Left ventricular end-diastolic internal diameter increased in all except the two switch failures. In all the successful cases the left ventricle had assumed a circular shape on cross-section with the interventricular septum contracting in synergy with the left ventricular mass. In the two failures, however, the interventricular septum had remained flat.
Conclusions. Echocardiography can be used reliably in surgical decision making in rapid two-stage arterial switch operation. Increase in left ventricular mass, left ventricular posterior wall thickness, and left ventricular end-diastolic internal diameter toward normal combined with an acquisition of circular left ventricular configuration with the interventricular septum contracting in synergy with the left ventricular mass appear to best predict successful outcome.
The arterial switch operation is the preferred mode of treatment for simple transposition of great arteries (intact ventricular septum or a small ventricular septal defect [VSD]) [1–4]. Progressive regression of left ventricular (LV) mass in simple transposition of great arteries (TGA) after birth [5–7] restricts the safe period for a primary arterial switch operation to the first month of life [8]. Left ventricular preparation by means of pulmonary artery banding and an aortopulmonary shunt followed by an arterial switch 5 months later was first reported by Yacoub and co-workers in 1977 [9]. The principal disadvantages with this approach are dilatation of the proximal main pulmonary artery, neoaortic valve regurgitation [10], and the need to interpose a prosthesis to bridge the gap in the newly constructed pulmonary artery [9]. The rapid two-stage arterial switch operation, first reported in 1989, overcame these disadvantages because of the short interval between the two stages [11]. In India, a sizeable number of patients with TGA reach hospital well beyond the first month of life because of inadequate health care systems. Before December 1991 all such patients underwent a Senning repair at our institution. Since December 1991, however, these patients were entered into a program for rapid two-stage arterial switch. Serial echocardiography was used exclusively to monitor LV preparation and for timing of the second stage operation. On the basis of this experience we have attempted to define echocardiographic criteria for assessment of adequacy of LV preparation for a successful second stage arterial switch operation.
Material and Methods
All patients with TGA with essentially intact ventricular septum more than 1 month of age were enrolled into a protocol for rapid two-stage arterial switch operation. Seventeen patients with ages ranging from 2 to 60 months (median age, 4 months) and body weights ranging from 3.5 to 11 kg (median, 4.0 kg) underwent the first stage procedure between December 1991 and September 1993. Fifteen patients had intact ventricular septum and 2 had a restrictive VSD. Late referral was the most common reason for delay in presentation. Balloon atrial septostomy was attempted in 14 of the 17 patients and was unsuccessful in 1 patient. Left ventricular preparedness was assessed by serial echocardiograms after the first stage operation. Cardiac catheterization between the preparatory and the definitive operation was performed only in 1 patient (patient 2).
Echocardiographic Evaluation
A baseline M-mode and two-dimensional echocardiogram was performed before the first stage operation and the left ventricular end-diastolic internal dimension (LVIDed), end-systolic internal dimension, posterior wall thickness (LVPWth), and interventricular septal thickness (IVSth) were measured according to the recommendations of the American Society of Echocardiography [12]. The left ventricular mass was derived by [(LVIDed + IVSth + LVPWth)3 - (LVIDed)3)] x 1.05. The subcostal and parasternal short axis views were used to determine the LV cross-sectional appearance and interventricular septal configuration. All patients underwent at least two echocardiographic evaluations after the first stage operation and the above mentioned parameters were recorded. In addition, the Doppler gradient across the pulmonary artery band was determined and attempts were made to detect shunt signals in the pulmonary artery distal to the band. Left ventricular measurements were also obtained from echocardiograms performed on 28 infants and children matched for age and body surface area. This group served as control population.
First Stage
A median sternotomy approach was used in all but 1 patient. The option of an atrial repair was always kept at this stage in case the coronary anatomy proved unsuitable for coronary transfer. A 4-mm polytetrafluoroethylene graft (5 mm in 1 patient) was placed between the base of the right subclavian or innominate artery and the right pulmonary artery. A band was then placed around the main pulmonary artery. In the initial experience LV pressure was monitored continuously perioperatively using a 19-gauge catheter placed retrogradely across the pulmonary valve in the LV. Subsequently, pressures were recorded intermittently from the proximal pulmonary artery or LV with a needle during banding. The band was tightened to elevate LV pressure to at least 75% of systemic pressure while maintaining a minimum systemic arterial saturation of 75%.
Interval Period or Period of Left Ventricular Hypertrophy
All patients received dopamine infusion (5 to 15 µg • kg-1 • min-2). Their clinical course in the intensive care unit after the first stage operation was often stormy and similar to the earlier report on this subject [13]. The first 24 hours were characterized by sinus tachycardia (with heart rates between 160 to 200 beats/min), low arterial pressure, metabolic acidosis, and oliguria nonresponsive to diuretics. Large volume infusions were required to maintain cardiac output with resultant visible swelling of the body. After the first 2 days the heart rate gradually normalized and there was improvement in the systemic blood pressure and peripheral perfusion. Spontaneous diuresis and improvement in systemic oxygenation followed. Eleven patients needed pressor support for more than 72 hours and 8 of these for more than 96 hours. Only 2 patients were on ventilator for less than 72 hours, and 4 needed to be ventilated for the entire duration of their intensive care unit stay (mean, 8.6 days) (Table 1
). Two of these 4 patients had a coexistent restrictive VSD. The combined volume loading of the left heart by the Blalock shunt and the VSD, together with left ventricular outflow obstruction produced by the band, produced significant pulmonary venous hypertension (confirmed on chest roentgenogram) that precluded successful extubation. The third patient (patient 7) became significantly hypoxic when weaning from the ventilator was attempted on the third postoperative day after the first stage. Echocardiography done on the third postoperative day suggested a blocked shunt, but it was decided not to redo the shunt as the LV seemed to be undergoing satisfactory hypertrophy and ventilator support was continued. The fourth patient showed persistent roentgenographic features of increased pulmonary blood flow and failed extubation on the third postoperative day. A loose band probably resulted in excessive pulmonary blood flow and she was maintained on ventilation until the second stage. Patient 6 had severe bradycardia immediately on transfer to the intensive care unit from the operating room after the first stage operation. He was rushed back to the operating room and the pulmonary artery band was removed. Continued hemodynamic instability, however, necessitated an emergency Senning repair, after which he made an uneventful recovery. The ratio of left ventricular to right ventricular systolic pressure had been 1.0 immediately after the first stage, and no reason for acute hemodynamic collapse could be ascertained. Patient 5 was discharged from hospital on the 15th day after the first stage as his left ventricle had not hypertrophied satisfactorily. He was readmitted on day 22 for progresive congestive heart failure. Echocardiography at this time showed that the LV was significantly dilated and it was decided to subject him to definitive repair.
|
-receptor blockade: (1) to facilitate uniform cooling and rewarming and (2) to ensure minimum systemic afterload to the newly prepared LV at the time of weaning from cardiopulmonary bypass. The pulmonary arteries were dissected up to the lung hila after division of the ductus arteriosus. The aorta was cross-clamped, cardioplegia given to the aortic root, and the aorta transected. Coronary buttons were harvested. The main pulmonary artery was transected 3 to 4 mm below the bifurcation. The coronary buttons were sutured to medially placed trap door flaps created in the main pulmonary artery stump. The Lecompte maneuvre was then performed and the aorta reconstructed. Under a short period of circulatory arrest the atrial septal defect and, when present, the VSD was closed. Pericardial patch reconstruction of the neopulmonary artery was carried out after release of aortic clamp during rewarming. The anterior part of the pulmonary artery was always a direct anastomosis between the residual portion of the proximal aortic stump and the distal pulmonary artery circumference. Appropriate monitoring lines were placed and the patient weaned from cardiopulmonary bypass. Results
There was no death after the first stage operation. One patient had to have emergency Senning repair because of acute hemodynamic collapse after the first stage. The remaining 16 patients progressed to an arterial switch operation. There was one mortality after the second stage operation. Thus overall mortality was 5.9%.
Patient 2 (aged 60 months; the oldest in this series) could not be weaned off cardiopulmonary bypass after arterial switch despite unequivocal evidence of good coronary perfusion. Successful salvage was achieved by immediate takedown of arterial switch and conversion to a Senning repair. Subsequently, he made an uneventful recovery.
Patient 7 (aged 7 months) was weaned off cardiopulmonary bypass with borderline pressures on considerable inotropic support. He had a downhill course in the intensive care unit and died 2 days after arterial switch with biventricular failure. Autopsy revealed patent coronaries. There was no evidence of myocardial infarction. Left ventricular wall thickness was adequate; however, the LV cavity was quite small.
Patient 11 had a VSD that was thought to be very small and therefore, was not closed during the arterial switch operation. Postoperatively, however, the patient did not tolerate extubation. Repeat cardiac catheterization showed a left-to-right shunt of 1.6:1 across the VSD. The patient was reoperated for closure of VSD after which she recovered uneventfully. Patient 15 developed complete heart block after the second stage operation which included direct closure of a small VSD and needed permanent pacemaker implantation.
Predischarge echocardiograms in the 14 switch survivors showed good LV function in all with no regional wall motion abnormalities. Two had normotensive pulmonary regurgitation. Four patients had gradients of less than 20 mm Hg across the pulmonary valve. None had any aortic incompetence.
Changes in Echocardiographic Parameters After the First Stage
All patients showed an increase in LV mass (from 40.8 ± 17.8 to 81.4 ± 25.4 g/m2) and LVPWth (from 3.3 ± 0.46 to 4.63 ± 0.58 mm (Fig 1; Table 2). The fractional shortening fell in all but 3 patients (in whom there was essentially no change) after the first stage operation. The mean pressure gradient across the pulmonary band was 45.8 ± 9.0 mm Hg. All but one patient (patient 7) had continuous shunt signals recorded in the pulmonary artery. The echocardiographic feature that characterized the two failed procedures was a flat interventricular septum in the cross-sectional view (Fig 2), before the second stage operation. An additional feature in these 2 patients was the absence of increase in LVIDed after the first stage (Fig 1C; Table 2).
|
|
|
). The LVIDed increased in all but 1 patient with successful second stage operation (Fig 1C). This patient (patient 15, aged 5 months) had a restrictive VSD and had a normal LVIDed before the first stage of operation. A two-stage procedure was performed instead of a single stage procedure because of a flat septal configuration and subsystemic LV pressures (left-to-right ventricular pressure ratio of 0.6).
Follow-Up
Follow-up ranges from 2 to 24 months (median, 12 months). All patients are doing well symptomatically. Echocardiograms repeated at 12 to 24 months follow-up in all patients shows absence of significant aortic valve incompetence and normal left ventricular function (mean LV ejection fraction of 0.80; range, 0.65 to 0.94).
Comment
Primary arterial switch in the first month of life is the preferred treatment for TGA [4]. For patients presenting beyond the neonatal period, anatomic correction can still be achieved, although in two stages [9]. Previously a successful arterial switch was presumed to be possible only after an interval of 6 weeks to 6 months after the preparatory stage [2]. This approach, however, is fraught with the disadvantages of band migration, proximal pulmonary artery distortion, and pulmonary annular dilatation resulting in neoaortic valve regurgitation after arterial switch, and the need for a prosthetic tube to bridge the gap in the neopulmonary artery [10, 14]. The demonstration of the rapidity of cardiac myosite hypertrophy in the biochemical laboratory [15, 16] followed by successful application of the concept to the clinical setting has resulted in the rapid two-stage arterial switch operation [11]. This approach has been reported to avoid the undesirable features of the delayed two-stage repair as the mean interval between the two stages can be as short as 7 days [1]. Our own experience confirms this finding. In the 14 patients who had successful completion of their two-stage repair, there was no incidence of early aortic valve incompetence. We were able to reconstruct the pulmonary artery just as for a primary arterial switch in all patients.
Because of increasing reports of attempts at retraining the LV and subsequent arterial switch for late right ventricular failure after arterial repair of TGA [17, 18] and the complexity and uncertainty of the operation at that stage [18], it seems only logical to perform the two-stage arterial switch at presentation rather than after a failed atrial repair. The cardiac myosite of the infant is capable of a hyperplastic response to pressure and volume loading [19]. As angiogenesis accompanies myosite proliferation up to 3 to 6 months of age, this is obviously a more physiologic type of accomodative response than one of pure hypertrophy that results only in increase in cytoplasmic volume of the myosite. Undoubtedly, the best time to prepare the LV of TGA for systemic afterload is in infancy.
What constitutes a prepared LV? We analyzed the clinical and echocardiographic data in all our patients to identify criteria that would ensure a safe arterial switch after a short preparatory interval. The following observations were made.
Age
Hypertrophy seemed to occur more slowly in the older age group. Although we were not able to establish a cut-off age beyond which a rapid two-stage approach will definitely not be feasible, we believe that at age of 1 year strict attention needs to be paid to LV dimensions before proceeding for a second stage repair. Two of our patients were more than 1 year in age. Patient 5 was 18 months old and had a left-to-right ventricular pressure ratio of 0.9 after band placement. However, as the LV mass had not increased satisfactorily by 15 days, he was discharged from hospital for a proposed delayed two stage repair. Progressive congestive heart failure, however, necessitated readmission and repeat echocardiographic examination on the 22nd day after the first stage. At that time, the LV seemed to be adequately hypertrophied (LVIDed 6.5 mm/m2 [baseline 5.1 mm/m2]; LVPWth 5 mm [baseline 3.5 mm]; and an increase in LV mass to 72.7 g/m2 [baseline 34.7 g/m2]). However, the fractional shortening had dropped markedly from 66% to 33%. At operation, pressures were taken. Demonstration of left-to-right ventricular pressure ratio of 2.0 gave us the confidence to proceed with an arterial switch operation. In this patient, the band gradient had suddenly increased in the third week of the preparatory phase probably because of distal migration of the band giving rise to the acute decompensation and, in all probability, the additional pressure stimulus had rapidly prepared the left ventricle.
The second patient (patient 2) was 60 months old and was the second case in our series. His band had been tightened to give a left-to-right ventricular pressure ratio of 0.8 along with a 5-mm Blalock shunt. At the end of the 2-week period, the LVPWth had increased from 4 to 5 mm and LV mass had increased from 67 g/m2 to 131.4 g/m2. However, there was no increase in LVIDed, and the interventricular septum was flat and seemed to move with the right ventricle. This patient was catheterized and found to have systemic LV pressure. Being unsure of the interpretation of the echocardiographic data early in our experience this patient was subjected to an arterial switch on day 16 after the first stage. The patient could not be weaned from cardiopulmonary bypass despite patent coronaries. The arterial switch was taken down and converted to a Senning repair. On retrospective analysis the impressive increase in LV mass along with equalization of left and right ventricular pressures proved misleading. Significantly, there had been no increase in the LVIDed.
Role of Atrial Septal Defect
Another aspect separate from the ability of the cardiac myosite to hypertrophy in a short time in patients beyond infancy is the presence of a large atrial communication in all patients of TGA with intact ventricular septum who survive to this late age. Although the presence of a sizeable atrial communication is desirable for the safe conduct of the preparatory operation (and hence the need for a balloon atrial septostomy before first stage) [13], a large decompressing atrial communication will restrict the volume loading of the left ventricle. To what extent this factor plays a role in patients of TGA beyond 1 year of age and whether this particular factor can be offset by a tighter band and construction of a larger shunt, is unclear.
Left Ventricular/Right Ventricular Pressure Ratio
A high left-to-right ventricular pressure ratio is definitely a basic requirement for induction of LV hypertrophy. This parameter, in isolation, is not an indication of LV preparedness. This can be deduced from the finding that patients with ratios 1.0 and 1.1 failed to have a successful arterial switch (patients 2 and 7), whereas others with ratios as low as 0.77 (patient 1) had a successful outcome (see Table 1). Therefore, we do not believe that cardiac catheterization and measurement of LV pressure is essential for assessing LV preparedness.
Echocardiographic Indices of Left Ventricular Preparedness
We have had a consistent significant increase in LV mass in all our patients including those who failed arterial switching (Fig 1A). Even doubling of LV mass did not guarantee an adequately prepared LV. However, calculation of LV mass by echocardiography before the first stage operation is prone to error owing to the banana-like shape of the LV that precludes most geometric assumptions. As the LV changes shape to spherical through the preparatory phase, LV mass calculations for the two shapes using the same formula may not be comparable. As such LV mass calculations are dependent on wall thickness and LVIDed. Hence, recently we have been relying more on LVIDed and LVPWth individually rather than on LV mass.
A LVPWth of as low as 3.4 mm and 3.7 mm resulted in a successful, uneventful arterial switch opertion (Patients 3 and 13). These results are in variance from an earlier report on the delayed arterial switch operation [20] where an LVPWth of 4 mm was the least admissable for a successful arterial switch.
That LV volume loading is of definite importance (and this has not been stressed in either of the earlier reports) [11, 20] is quite clear. Patient 7, who died after arterial switch, had a satisfactory LV hypertrophy in terms of LVPWth (increase from 3.5 to 5.5 mm) and LV mass (increase from 19.5 to 39.7 g/m2 compared to 46.0 g/m2, which is the minimum LV mass in the age, height, weight-matched control population), with a left-to-right ventricular pressure of 1.1. There was, however, no increase in LVIDed (4.3 mm/m2 compared to 7.2 mm/m2, which is the minimum LVIDed in the age, height, and weight-matched control population). This could have been a result of the blocked Blalock shunt that failed to provide the LV with the required preload. The other patient whose LVIDed did not increase (patient 2) also failed arterial switching (discussed above). The only patient who had a successful arterial switch despite no increase in LVIDed (patient 15) also had a restrictive VSD that had probably already volume loaded the LV adequately (LVIDed 7.9 mm/m2) before the first stage.
Left ventricular fractional shortening either remained the same or dropped (sometimes significantly), and in 1 patient (patient 5) drastically. Change in LV fractional shortening after the first stage did not distinguish the patients who tolerated arterial switching from those who did not.
Notwithstanding the numerical values of the above mentioned parameters, we have relied significantly on the visual appearance of the LV and interventricular septum on the cross-sectional echocardiogram. Specifically a circular cross-sectional appearance of LV in the short axis view appears to best indicate LV preparedness for a second stage operation (Fig 3). This would require the interventricular septum to become convex toward the right ventricle after the first stage instead of being convex toward the LV as it usually is in a patient with simple TGA beyond the neonatal period. In this context the results of a recent report of single stage arterial switch operation in infants older than 21 days is of particular interest [21]. In this study the patients were classified on the basis of echocardiography as an unequivocally low (banana-shaped LV), probably low (bowing of septum toward left ventricle), and systemic (normal LV geometry) LV pressures. Cardiac catheterization data did not correlate well with the echocardiographic appearance. Despite this, the researchers used the echocardiographic findings in their final analysis and found that patients classified on echocardiography as having unequivocally low or probably low LV pressures had longer intensive care unit stays and required longer periods of inotropic support indicating that the cross-sectional LV configuration is a useful indicator of LV preparation. Our experience is quite consistent with these observations.
|
Footnotes
Presented at the Sixth International Symposium on Echocardiography in Cardiac Surgery, Washington, DC, Nov 9–11, 1994. Received the Doppler Award for best paper presented at the symposium.
Address reprint requests to Dr Iyer, Department of Cardiothoracic & Vascular Surgery, Cardiothoracic Sciences Centre, All India Institute of Medical Sciences, Ansari Nagar, New Delhi-110 029, India.
References
This article has been cited by other articles:
|
|
 |
|
  R K Kumar and S Shrivastava Paediatric heart care in India Heart, August 1, 2008; 94(8): 984 - 990. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C.D. Abduch, R. S. Assad, M. Q. Rodriguez, A. S. Valente, J. L. Andrade, L. M.M. Demarchi, M. B. Marcial, and V. D. Aiello Reversible pulmonary trunk banding III: Assessment of myocardial adaptive mechanisms--contribution of cell proliferation J. Thorac. Cardiovasc. Surg., June 1, 2007; 133(6): 1510 - 1516. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Sivakumar, E. Francis, P. Krishnan, and J. Shahani Ductal stenting retrains the left ventricle in transposition of great arteries with intact ventricular septum. J. Thorac. Cardiovasc. Surg., November 1, 2006; 132(5): 1081 - 1086. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Sharma, S. K. Choudhary, A. Bhan, R. Juneja, S. S. Kothari, A. Saxena, and P. Venugopal Left ventricle is better suited as pulmonary ventricle in simple transposition with severe pulmonary hypertension Ann. Thorac. Surg., November 1, 2002; 74(5): 1612 - 1615. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Sharma, A. Bhan, R. Juneja, S. Sunder Kothari, A. Saxena, and P. Venugopal Double switch for congenitally corrected transposition of the great arteries Eur. J. Cardiothorac. Surg., March 1, 1999; 15(3): 276 - 282. [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 |
