Total Correction of Transposition of the Great Arteries in Infancy as Initial Surgical Management

Transcription

1 Total Correction of Transposition of the Great Arteries in Infancy as Initial Surgical Management Lawrence I. Bonchek, M.D., and Albert Starr, M.D. ABSTRACT Fifteen patients have undergone total correction of transposition of the great arteries (TGA) in the past two years. Four were under 6 months of age and weighed 3.4, 3.5, 4.3, and 5.7 kg., respectively; 6 were from 13 to 21 months old; and 5 were 2 to 10 years old. In this series, balloon atrioseptostomy (BAS) provided palliation that was brief and often unsatisfactory. All 8 patients who had BAS without surgical septectomy had severe cyanosis, exercise intolerance, and retardation of growth and development prior to correction. There were 2 operative deaths related to complicated TGA and 1 unrelated late death. All survivors have improved dramatically. Survival alone is therefore no longer considered to be successful palliation following BAS. If adequate relief of cyanosis and reasonable growth and development are not achieved, total correction in infancy is a favorable alternative to surgical septectomy. D espite success with the Mustard technique for correction of transposition of the great arteries (TGA) in older children, most centers prefer palliative procedures in infants. The Mayo Clinic recently reported a 5% mortality for the correction of uncomplicated TGA in childhood and a 40% mortality in infancy [2]. The palliation afforded by balloon atrioseptostomy (BAS), however, is often brief and of unsatisfactory quality. Cyanosis, exercise intolerance, and marked retardation of growth and development usually persist, and cerebrovascular accidents can occur. Thus, surgical intervention often becomes necessary in infants and atrial septectomy is preferred, but the mortality is as high as 20 to 25% [4]. For these reasons, the Mustard procedure is now performed at the University of Oregon Medical School Hospital whenever BAS fails to provide satisfactory palliation, regardless of the patient s age. This report reviews our experience with correction of TGA, describes recently adopted features of operative technique and postoperative management, and emphasizes results with this approach in the past two years. From the Department of Surgery, University of Oregon Medical School Hospital, Portland, Ore. Supported in part by National Institutes of Health Career Development Award HE to Dr. Bonchek. Presented at the Eighth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, Calif., Jan , Address reprint requests to Dr. Bonchek, University of Oregon Medical School Hospital, 3181 S.W. Sam Jackson Park Rd., Portland, Ore THE ANNALS OF THORACIC SURGERY

2 Total Correction of Transposition in Znfancy Clinical Material Twenty-one patients with TGA have had corrective operations at the University of Oregon Medical School. Six patients provided preliminary experience with the Mustard procedure prior to 1970, and 15 patients have undergone operation in the past two years with modified operative techniques. In the initial group, 4 patients were under 2 years old (2, 11, 19, and 22 months, respectively) and 2 were between 4 and 5 years of age. One infant had a large patent ductus arteriosus (PDA), another had PDA and subpulmonic stenosis, and 2 had uncomplicated TGA. One older child had a trivial ventricular septa1 defect (VSD), and the other had uncomplicated TGA. Only the 3 oldest children had undergone surgical septectomy. Two of the 4 infants died postoperatively, 1 of profound pulmonary insufficiency from an unrecognized PDA and the other from heart block and ventricular irritability. One infant died suddenly at home 4 months postoperatively, two weeks after a routine clinic visit; a postmortem examination was not obtained. The 4-year-old died eleven days postoperatively with pneumonia and hypoxia following massive intratracheal hemorrhage from the site of a tracheostomy done on the fifth postoperative day. He had severe clotting abnormalities and twice had required reexploration for intrathoracic bleeding immediately postoperatively. Although overall results were poor, the initial group of patients had demonstrated the technical feasibility of corrective operation in infancy using cardiopulmonary bypass. Operative techniques and postoperative routines were reviewed and revised, and the current program was initiated. The current group consists of 10 infants and 5 older children who have undergone corrective operations in the past two years. Of the 10 infants less than 2 years old, 4 were under 6 months and weighed 3.4, 3.5, 4.3, and 5.7 kg., respectively. Six patients between 12% and 21 months old weighed from 6.4 to 10.1 kg. (Table). Five infants had no VSD or pulmonic stenosis (PS), 1 had a large VSD with elevated pulmonary vascular resistance (15 Wood units), 1 had a VSD and severe subpulmonic stenosis, 1 had a small VSD, and 2 had subpulmonic stenosis. Eight infants had had no prior procedures other than cardiac catheterization and BAS; 2 had had surgical septectomy. In 7 infants with BAS alone, operation was urgently indicated because of marked growth retardation, profound cyanosis, and severe chronic heart failure. In an eighth infant, severe cyanosis and a high hematocrit (80%) despite surgical septectomy at 1 month of age aroused concern that a cerebrovascular accident was imminent. In 2 infants the operation was electively performed at 15 and 16 months of age, respectively. None of the 5 older patients between 27 months and 10 years of age had a VSD or significant PS. (One had a left ventricular/pulmonary artery VOL. 14, NO. 4, OCTOBER,

3 TOTAL CORRECTION OF TRANSPOSITION IN 15 PATIENTS (10 INFANTS), Preop. Studies Age Weight BAS Septectomy Hct. 02Sat. Patient (yr.) (kg.) Anatomy (age) (age) (%) (%) Remarks B. B. 12 days 3.5 Large PDA 1 day None Emergency operation; died from S. R. D. V. B. K. G. M. T. T. M. B. 7 wk. 942 K2 1x2 lf Small VSD Large VSD Uncomplicated Subpulmonic stenosis VSD, subpulmonic stenosis Uncomplicated Unsuccessful None 2 days 1 wk. (x2) 1 wk. 1 wk complications related to ligation of PDA None Weight at operation was below birth weight; no resistance to withdrawal of 2 ml. balloon at BAS; no postoperative complications None None 1 mo. None Urgent operation for CHF; no complications Required diuretics for 10 days postoperatively Urgent operation for high hematocrit; no complications; subpulmonic stenosis developed after 1 week of age Died in operating room of low cardiac output with high LV pressure None Postoperative SVC obstruction; successful reoperation; late death from allergic bronchitis with pneumonia

4 K. D. 1? Uncompli- 3 wk. cated 4 wk. L. J. 1 Kz 6.6 Uncompli- 1 day cated W. B. 1% 6.4 Subpulmonic 1 wk. stenosis 2 wk. G. W D. D CI D. P. 4%~ 15.6 ce % 0 cp T. M None None 8 mo. Uncompli- 1 wk. None cated 5 mo. Mild subpul- 2% mo. 3 mo. 57 monic stenosis Mild PS 1 wk. 2 wk. 65 Uncompli- None 4 days 52 cated Uncompli- None 3 mo. 53 cated Elective operation; no complications Operation for severe growth retardation; no complications Operation for severe growth retardation; unrelieved subpulmonic stenosis; required prolonged respiratory support 56 Operation for growth retarda- tion; no complications 60 Operation for persistent cyanosis; no complications 70 Elective operation; no complications 85 Elective operation; no complications 85 Reoperation for bleeding Hct. = hematocrit; 0, Sat. = oxygen saturation; PDA = patent ductus arteriosus; VSD =ventricular septa1 defect; PS = pulmonic stenosis; BAS = F: balloon atrioseptostomy; CHF = congestive heart failure; LV = left ventricle. 9 w W

5 BONCHEKANDSTARR gradient of 36 mm. Hg, and relief of this mild obstruction was unnecessary.) All but 1 of these older children had had surgical septectomy in infancy prior to elective correction. One patient required correction at 27 months of age for increasing cyanosis and underdevelopment following BAS without prior septectomy. Operative Met hods ANESTHESIA Anesthesia is induced and maintained with halothane, succinylcholine, and curare. The halothane is administered through a nasotracheal tube in a nonrebreathing system. Arterial pressure is monitored with a Doppler ultrasonic transcutaneous flow detector" placed over the radial or brachial artery. (Arterial cut-downs are no longer used.) Postoperatively, all patients are returned to the cardiac surgical recovery suite with the nasotracheal tube in place regardless of their clinical condition. It is important that anesthetic management in infants be consistently supervised by the same experienced individual. OPERATIVE TECHNIQUES Operation is performed at normothermia with high-flow cardiopulmonary bypass (2.4 liters per minute per square meter of body surface area) utilizing a disposable infant oxygenatort primed with fresh heparinized blood. The cavae are cannulated at their junctions with the atrium as far to the patient's right as possible so that the atriotomy can be made to the left of the catheters to facilitate exposure. Electrical fibrillation is induced and maintained with the same wire that is used for ventricular pacing postoperatively, and no ventricular vents are used. The atrial baffle is made from a segment of pericardium (stretch-knit Dacron was used in 1 patient) cut in a generous rectangle with no preliminary sculpturing. This provides a baffle that is always adequate in circumference and that can be accurately and simply tailored during insertion if necessary, as described below. Suturing is begun in the usual manner at the left edge of the orifices of the left pulmonary veins (Fig. l) and proceeds to the right and superiorly to exclude the superior caval catheter (Fig. 2). Similarly, a suture is carried to the right and inferiorly to exclude the inferior caval catheter and the coronary sinus. Sutures are placed superficially around the coronary sinus, and it has been completely unnecessary to employ incisions of the sinus or the Aberdeen flap, as there have been no instances of heart block. Intermittent aortic cross-clamping is used to facilitate exposure if needed and has been used mainly while suturing in the vicinity of the coronary sinus. When 1 or 2 cm. of baffle remain to be sutured along the interatrial septa1 'Parks Laboratories Inc., Beaverton, Ore. ttemptro1, Bentley Laboratories, Inc., Santa Ana, Calif THE ANNALS OF THORACIC SURGERY

6 Total Correction of Transposition in Znfancy 2 FIG. 1. The rectangular pericardial patch now used for correction of TGA without preliminary sculpturing. FIG. 2. Caual cannula placement is done to the patient s right. This facilitates intmcardiac exposure considerably. The suture line has now excluded the superior caual cannula. 3 FIG. 3. Prior to completion of the suture line, a vertical incision (dotted line) can increase patch circumference if necessary. remnant between the mitral and tricuspid valves, the free edge of the baffle is adjusted as necessary. If the circumference of the baffle is inadequate to complete the suture line, an incision is made perpendicular to the free edge to provide the extra margin (Fig. 3). If there is excess free margin, the dogear is excised and the cleft so created is closed perpendicular to the free edge (Fig. 4). Suturing of the baffle is then completed, and the baffle fills with coronary sinus return; but if this is inadequate to fill the,baffle, a caval tourniquet can be loosened. In most cases, the baffie balloons upward in its midportion. To avoid pulmonary venous obstruction, the redundancy is reefed with a horizontal mattress suture (Fig. 5) carried back upon itself as an over-and-over suture for 1 to 2 cm. (Fig. 6). To avoid constriction, this suture must not reach the vestibule around either caval orifice. This technique has made it unnecessary to enlarge the new physiological left atrium in any patient in this series, and there have been no instances of VOL. 14, NO. 4, OCTOBER,

7 BONCHEKANDSTARR FIG. 4. Excess free margin can be excised as a dog-ear and the resultant cleft closed zoilh a running suture. pulmonary venous obstruction. Excision of a portion of the patch at this point is avoided since any incision in the baffle will collapse it, make orientation more difficult, and allow coronary sinus return to obscure vision unless the aorta is cross-clamped. The tailoring maneuver is more accurate with a distended bae. Furthermore, the horizontal mattress suture can be easily removed or altered prior to completion of the over-and-over suture, whereas excision of a portion of the patch is an irrevocable procedure. When it has been necessary to incise the baffle perpendicularly to lengthen the free margin (as described above), the midportion of the baffle is automatically flattened and reefing is often unnecessary. / /! 5 6 FIGS. 5, 6. The tendency of the ba@e to balloon upward is corrected, if necessary, by use of a reefing suture. 382 THE ANNALS OF THORACIC SURGERY

8 Total Correction of Transposition in Znfancy The ventricular septa1 defects were approached through a short longitudinal right ventriculotomy and closed by Dacron felt patch or direct suture. Adequate exposure and safe, complete closure were given priority, and exposure through the atrium was poor. Post o p era t iue Man age m en t As noted earlier, the nasotracheal tube is retained initially in all patients. Respiratory assistance is provided with a Bourns respirator for infants under 10 kg. and with a Bennett MA-1 respirator for larger children. Extubation is accomplished within one to five days after a trial of spontaneous respiration with the nasotracheal tube in place. This is occasionally impractical in small infants, as the smaller nasotracheal tubes provide sufficient resistance to air flow to increase the work of breathing. If this is suspected, the infant is extubated, with prompt reintubation if hypoxemia, bradycardia, labored respirations, or respiratory acidosis develop. Management of respirators, extubation, and reintubation are supervised and carried out by the cardiac surgical service exclusively. With this approach, only one tracheostomy has been performed in the entire series. Frequent, adept tracheal suctioning under direct vision with a laryngoscope is essential after extubation and obviates many tracheostomies. The initial fluid allowance is 50 ml. per kilogram per 24 hours with electrolyte supplementation as indicated: this allowance is gradually increased as tolerated. With this stringent restriction, fluid retention has occurred in only 1 patient, who required diuretics for ten days postoperatively. Aldosterone antagonists have not been used. Temporary atrial and ventricular pacing wires are inserted in all patients, and management of the arrhythmias that commonly occur is greatly facilitated. Ten of the 15 patients had an intermittent or persistent nodal rhythm at a satisfactory rate on return from the operating room, but this invariably reverted to sinus rhythm before discharge. There was no clinically observable hemodynamic difference between sinus and nodal rhythm at comparable rates. The unusual configuration and lack of contractility of the new atrial septum may well impair the effectiveness of atrial systole following correction of TGA. R esu 1 ts Of the 10 patients in the infant group there were 2 hospital deaths in patients with complicated TGA and 1 late death unrelated to the operation. One hospital death occurred in a 12-day-old infant with a very large PDA. An encircling ligature cut through the PDA while it was being tied through a median sternotomy. A large amount of acid-citrate-dextrose blood was administered during efforts to control the resulting hemorrhage; irreversible VOL. 14, NO. 4, OCTOBER,

9 BONCHEKANDSTARR metabolic acidosis and myocardial depression ensued, and the patient died 24 hours postoperatively in severe acidosis with a low output. Another infant with a VSD and severe subpulmonic stenosis (left ventricular pressure 85 mm. Hg) could not sustain an adequate cardiac output after an uneventful bypass. Efforts to expose the subpulmonic area through the pulmonary artery during bypass had been unsuccessful. At postmortem examination the VSD repair and the interatrial baffle were intact, and severe muscular subpulmonic stenosis was evident. One patient developed evidence of superior vena caval obstruction with marked facial swelling and progressive head enlargement from 2 to 4 months postoperatively, and an angiogram revealed almost complete obstruction of superior vena cava (SVC) inflow. At reoperation the baffle was found to be constricted just below the entry of the SVC into the new right atrium, as in the patient reported by Mazzei and Mulder [5]. This area was widened with a synthetic patch, and the SVC pressure declined from 30 mm. Hg to 14 mm. Hg. He was subsequently discharged doing well hemodynamically. However, he had been troubled since birth with numerous allergies and asthmatic bronchitis. His pulmonary problems persisted following both operations, and he was readmitted 1 month later with severe bilateral pneumonia from which he died. At postmortem examination the patch was found to be well healed, and pulmonary venous obstruction was clearly excluded. The oldest infant (21 months) required a tracheostomy and intermittent ventilatory assistance for almost 3 months. He had residual subpulmonic obstruction and marked cardiomegaly, and a recurrent left lower lobe collapse was attributed to this. His development had been severely retarded preoperatively, and he weighed 6.4 kg. at 21 months. He was discharged nearly 4 months postoperatively and is now gaining weight. The only complication in the group of 5 older patients was in the 10- year-old boy who required reoperation for bleeding and subsequently did well. Comment Balloon atrioseptostomy was introduced by Rashkind in 1966 and provides a relatively safe initial palliative procedure for TGA. Unfortunately, the initial improvement is often mild and short-lived. In a closely followed series of 43 patients reported by Baker and his associates [l], 31 were initially thought to have had a successful BAS. Eleven subsequently required surgical septectomy, and 9 of these had had two attempts at BAS. Seven others had a VSD without PS, which probably improved intracardiac mixing. Thus, in only 13 of 43 patients was effective palliation attributable to BAS alone. Others have reported similar experiences with BAS [4]. Especially noteworthy is the report by Waldhausen and associates [9] of 18 patients with uncomplicated TGA who had had a BAS performed by Dr. Rashkind. Four re- 384 THE ANNALS OF THORACIC SURGERY

10 Total Correction of Tj-amposition in Znfancy quired surgical septectomy, and 3 others had cerebrovascular accidents preoperatively. At the time of total correction of TGA in infants who have undergone BAS at the University of Oregon, it has been common to find stretching of the foramen ovale with no distinct tear and no disruption of the septum. In this series, BAS did not consistently result in prolonged palliation, and the quality of survival was inferior to that provided by surgical septectomy. Surgical septectomy, however, though often dramatically effective for long periods, has a high mortality (20 to 25% in many reported series and in our own) [4]. These disappointing experiences resulted in our recent policy of total correction of TGA in infants who require operation. Experience with the current group indicates the feasibility of correction of TGA in infancy using cardiopulmonary bypass. This is substantiated by our experience with other lesions, including tetralogy of Fallot, as reported elsewhere [6a]. Improvement in growth and development has been striking (Fig. 7), unlike the result with palliative procedures. Closure of isolated VSD has been uneventful, but management of associated severe subpulmonic stenosis remains unsatisfactory. Complete relief of subpulmonic stenosis in infants with TGA may be impossible in certain cases because of the length and the diffuse, muscular nature of the outflow tract obstruction. Plauth and his co-workers [6] on the basis of serial hemodynamic studies have suggested that subpulmonic obstruction may develop in patients with TGA and an intact ventricular septum. Their conclusions were only inferred, since the pulmonary artery was never entered on their initial studies. In 1 of our patients (G. M., Table), however, the development of PS was clearly documented. The left ventricular pressure was 35 mm. Hg at one week, and intraoperative pressures at 13 months of age showed a left ventricular pressure of 75 mm. Hg with a FIG. 7. Preoperative and 6-month postoperative 90 weights of the 8 youngest survivors expressed as percentiles from standard growth charts. W. B., L. J., 80 and G. M. were far below the third percentile preoperatively, and their substantial weight gain was not reflected in significant percentile changes 6 months postoperatively. U 60 2 e C, t 40 zl PREOP POSTOP (6 mos.) VOL. 4, NO. 4, OCTOBER,

11 BONCHEKANDSTARR pulmonary arterial pressure of 20 mm. Hg. At catheterization 7 months postoperatively the pulmonary artery was entered, and the left ventricular/pulmonary artery gradient was found to be unchanged. If this sequence occurs because of the increased pulmonary blood flow that is almost invariably present in TGA, early correction may prevent such a course. Even when obstruction is present and cannot be relieved, the Mustard procedure may provide better palliation than alternative methods, as suggested by Lindesmith and his colleagues [3]. The Rastelli procedure may be indicated in older children with PS and a VSD [ZI. The effect of early correction on the development of increased pulmonary vascular resistance is also of interest. Viles, Ongley, and Titus [8] have found pulmonary vascular disease at postmortem examination in infants with TGA even under 6 months of age. An intact ventricular septum, PS, or pulmonary artery banding in infancy did not invariably protect the pulmonary vasculature. Thus, even when BAS is successful, pulmonary vascular changes may continue to progress. Stark and associates [71, however, in a clinical series of 33 patients with TGA and VSD undergoing pulmonary artery banding, found irreversible pulmonary arterial hypertension only in those over 1 year of age. Waldhausen and his colleagues [9] found no increase in pulmonary vascular resistance in 18 patients with uncomplicated TGA up to 9% years of age. If pulmonary artery banding in infancy can prevent pulmonary vascular injury, total correction of TGA with VSD should be even more effective. Certain infants with TGA may have pulmonary vascular changes from birth, however, and will not benefit as much from total correction. Since this problem is not yet resolved and since differentiation of patients in these groups is difficult preoperatively, it seems reasonable to eliminate increased pulmonary blood flow by total correction whenever possible. The clinical course of the infant in this series with a large VSD has been entirely uncomplicated, and catheterization 5 months postoperatively revealed a left ventricular pressure (pulmonic ventricle) of 40 mm. Hg in contrast to a preoperative pressure of 100 mm. Hg. Conclusions The prompt resumption of normal growth and development after correction of TGA contrasts markedly with the course of infants who are followed for prolonged periods after BAS. The low mortality of total correction in infancy suggests that survival alone should not be deemed successful palliation following BAS, but that adequate relief of cyanosis and congestive heart failure as well as reasonable growth and development should be sought, just as they are following surgical septectomy. If palliation is inadequate by these criteria, total correction in infancy provides a satisfactory alternative to surgical septectomy. 386 THE ANNALS OF THORACIC SURGERY

12 Total Correction of Tran.sposition in Infancy Addendum Since this manuscript was submitted, 2 additional patients have developed symptomatic superior vena caval obstruction at 4 months and 1 year postoperatively, respectively, despite the construction of a wide conduit for SVC blood at the first operation. In 1 patient a reefing suture had been utilized; in the other, no reefing suture was used. At reoperation, in both patients the obstruction was found to be clearly due to shrinkage and marked thickening of the pericardium. It seems apparent that in these very small infants, growth of the free wall of the heart did not keep pace with shrinkage of the pericardium. We now prefer stretch-knit Dacron for correction of TGA. References 1. Baker, F., Baker, L., Zoltun, R., and Zuberbuhler, J. R. Effectiveness of the Rashkind procedure in transposition of the great arteries in infants. Circulation 43 (Suppl. I):l, Ching, E., DuShane, J. W., McGoon, D. C., and Danielson, G. K. Total correction of cardiac anomalies in infancy using extracorporeal circulation. J. Thorac. Cardiovasc. Surg. 62:117, Lindesmith, G. G., Stiles, Q. R., Tucker, B. L., Gallaher, M. E., Stanton, R. E., and Meyer, B. W. The Mustard operation as a palliative procedure. J. Thorac. Cardiovasc. Surg. 63:75, Litwin, S. B., Plauth, W. H., Jones, J. E., and Bernhard, W. F. Appraisal of surgical atrial septectomy for transposition of the great arteries. Circulation 43 (Suppl. 1):7, Mazzei, E. A., and Mulder, D. G. Superior vena cava syndrome following complete correction (Mustard repair) of transposition of the great vessels. Ann. Thorac. Surg. 11:243, Plauth, W. H., Nadas, A. S., Bernhard, W. F., and Fyler, D. C. Changing hemodynamics in patients with transposition of the great arteries. Circulation 42:131, a. Starr, A., Bonchek, L. I., and Sunderland, C. 0. Total correction of tetralogy of Fallot in infancy. J. Thorac. Cardiovasc. Surg. In press, January, Stark, J., Tyran, M., Tatooles, C. J., Aberdeen, E., and Waterston, D. J. Banding of the pulmonary artery for transposition of the great arteries and ventricular septa1 defect. Circulation 42 (Suppl. 11): 116, Viles, P. H., Ongley, P. A., and Titus, J. L. The spectrum of pulmonary vascular disease in transposition of the great arteries. Circulation 40:31, Waldhausen, J. A., Pierce, W. S., Park, C. D., Rashkind, W. J., and Friedman, s. Physiolo@c correction of transposition of the great arteries. Circulation 43:738, Discussion DR. BERT W. MEYER (LOS Angeles, Calif.): The operative management of TGA in the young infant under 1 year of age has been convincingly presented. We approach the problem in a similar manner but with some significant exceptions. The cardiology department at Children s Hospital in Los Angeles has been a pioneer in the early diagnosis and appropriate treatment of cardiac defects in the young infant, but in TGA we favor initial palliation by opening the atrial septum by BAS and then, if this is not adequate, septectomy under direct VOL. 14, NO. 4, OCTOBER,

13 BONCHEKAND STARR vision with inflow occlusion. In our hands, clinical improvement has been good, and the mortality is less than 3%. The patient is subsequently followed medically until the age of 2 years, at which time the Mustard operation is done. In the young infant weighing less than 4 kg., bypass is more critical and intracardiac suturing is more difficult because of the small heart size. Also, postoperative management with regard to such factors as fluid balance, acid-base status, potassium level, and temperature control can be most trying. In our series of 46 patients, 3 were under 1 year of age and 2 of these patients died. There were only 4 deaths in all. Only 1 patient has died while on medical management following palliative septectomy. Pulmonary vascular changes and acquired subpulmonic obstruction may possibly be prevented by an early definitive operation; but this is indeed questionable, and we do not believe it is a valid reason to perform an early operation. Our surgical technique is similar to that described. Placement of the caval catheters is extremely important, and suturing of the baffle around the cavae must be done with care to avoid postoperative caval obstruction. A minor point is that we sew in the baffle exactly opposite to the way described; we place the long axis of the pericardial rectangle across the pulmonary veins. Tailoring the baffle then is rarely required. We do not fibrillate the heart, but the tricuspid valve is held incompetent at all times. The use of pacemaker wires as described by the authors probably has merit, but we do not use them. There have been no permanent heart blocks and no serious rhythm disturbances postoperatively. Exposure of the VSD through the systemic ventricle has given no problem, but we agree wholeheartedly that relief of obstruction at the pulmonic or subpulmonic level can be most difficult because of the inaccessibility of the area. Palliation of complicated TGA by the Mustard operation, as previously described by Dr. Lindesmith, warrants continued consideration. We are in complete agreement with the postoperative management described. The use of a nasotracheal tube with assisted ventilation postoperatively is vital to a successful result, and our anesthesia department has been most helpful in this regard. DR. JAMES W. KILMAN (Columbus, Ohio): I agree totally with Dr. Bonchek and Dr. Starr in their tenet concerning early repair of TGA. We too were dissatisfied with our palliative procedures for TGA and at the gradual failure of the medical BAS. Early last spring we assumed a more aggressive approach for small infants, believing that we could do better with total repair. We have now repaired defects in 12 infants, all under 10 kg. in body weight. The mean weight has been 8.9 kg. All of them had had at least one medical BAS, and 1 infant had had more than one. The preoperative poz has averaged 42 mm. Hg, and the postoperative poz on room air has averaged 76 mm. Hg. Our main complications have been arrhythmias. We put pacemaking wires in all these patients. All have been discharged from the hospital with no other serious complications, and we have seen an extremely low incidence of residual arrhythmias. We have had no deaths in this group of patients. DR. GEORGE A. TRUSLER (Toronto, Ont., Canada): In this beautiful presentation, Dr. Bonchek and Dr. Starr have indicated that children with TGA can have correction early in life with a low risk. Whether to operate early, say in the first year of life, or later must be decided by each individual surgeon. In a recent experience of one individual, correcting TGA in children over 12 months old is associated with a mortality that in total, including debandings, is approximately lo%, although in uncomplicated TGA the mortality is relatively ]OW. However, for comparison we must add the number of infants who, following BAS, fail to survive their first year. In our hospital this amounts to 12%. These infants may die at any time during that first year and without particular warning. 388 THE ANNALS OF THORACIC SURGERY

14 Total Correction of Transposition in Infancy Obviously, if the authors and others can achieve the low mortality presented today, we must examine our results and possibly lower the elective age at which we correct the malformation in order to avoid this natural attrition. In light of this, 1 would ask Dr. Bonchek what he now considers to be the elective age for correction in an infant with an uncomplicated form of TGA. DR. BONCHEK: Dr. Meyer, we certainly envy your results with palliative procedures. The fact that we could not duplicate such results with palliative operations led us to the policy of early total correction. Dr. Kilman, nodal rhythms have also been very common in our patients postoperatively, but the patients have invariably been in sinus rhythm at the time of discharge and we have not seen recurrent nodal rhythms or heart block in any of them. Dr. Trusler, the point we are trying to make is that BAS, though it is often lifesaving, provides palliation which is of very poor quality in many cases, and surgical intervention frequently comes under deliberation. It should be considered early if the patient has intractable heart failure, marked retardation of growth and development, or severe cyanosis; in such patients, total correction of TGA can be carried out at almost any age.