Early Reduction of the Volume Work of the Single Ventricle: The Hemi-Fontan Operation

Transcription

1 Early Reduction of the Volume Work of the Single Ventricle: The Hemi-Fontan Operation Marshall L. Jacobs, MD, Jack Rychik, MD, Jonathan J. Rome, MD, Sotiria Apostolopoulou, MD, Christian Pizarro, MD, John D. Murphy, MD, and William I. Norwood, Jr, MD, PhD Division of Cardiac Surgery, Department of Surgery, Deborah Heart and Lung Center, Browns Mills, New Jersey; and Division of Cardiovascular Surgery, Department of Surgery, and Division of Cardiology, Department of Pediatrics, The Children's Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania Background. In hearts with a functional single ventricle, cavity volume and myocardial muscle mass increase as a consequence of the excessive volume work associated with parallel pulmonary and systemic circulations. The hemi-fontan operation was conceived as a means of accomplishing early reduction of the volume work of the single ventricle. Methods. All patients presenting in infancy with single-ventricle physiology were managed by early hemi- Fontan operation in anticipation of a subsequent completion Fontan operation. Between May 1989 and August 1995, 4 patients less than 2 years of age underwent hemi-fontan operations. Mean age at operation was 8.5 months (range, 2 months to 24 months). The hemi-fontan operation included association of superior vena(e) cava(e) with the branch pulmonary arteries, augmenta- tion of the central pulmonary arteries, occlusion of the inflow of the superior vena cava into the right atrium, and elimination of other sources of pulmonary blood flow. Results. Operative mortality (<3 days) was 31 of 4 patients (7.8%). For the last 2 patients, operative mortality was 8 of 2 (4.%). Younger age at operation was not an independent risk factor for operative mortality. Urgent operation in the presence of a hemodynamic burden requiring concomitant procedures was associated with increased mortality. Conclusions. The hemi-fontan operation can be accomplished with low operative mortality in young patients, achieving early reduction of the volume work of the single-ventricle heart. (Ann Thorac Surg 1996;62:456-62) ontemporary management of patients with hearts C with a functional single ventricle involves eventual performance of some modification of the Fontan operation. The goals of the Fontan operation include reduction to normal of the volume work of the single ventricle, and achievement of normal or near-normal systemic arterial oxygen saturation. Before the Fontan operation is accomplished, the unoperated or palliated state is characterized by a single ventricular pump that simultaneously ejects blood into both the pulmonary and systemic circulations, and thus of necessity does volume work in excess of normal. The adaptive response to this abnormal volume load includes increases in both cavity volume and myocardial muscle mass. As is true in relation to volume overload of the systemic ventricle in the normal or four-chambered heart, the duration of the hemodynamic burden and the magnitude of the adaptive changes in the properties of the myocardium itself both have an impact on the outcome of a surgical procedure that removes the volume load [1]. Thus in the case of single-ventricle hearts, older age and increased ventricular mass have both been identified as risk factors for mortality in Presented at the Thirty-second Annual Meeting of The Socie~ of Thoracic Surgeons, Orlando, FL, Jan 29-3l, Address reprint requests to Dr Jacobs, Department of Surgery, Deborah Heart and Lung Center, 2 Trenton Rd, Browns Mills, N] 815. association with the Fontan operation [2]. Above and beyond the objective of minimizing operative mortality, preservation of ventricular function to optimize functional capacity and longevity of survival after the Fontan operation must be the goals of any protocol for the management of patients with single-ventricle hearts. Although reduction of the volume work of the single ventricle is a principal objective of the Fontan operation, it must be recognized that after removal of a chronic volume load, regression of myocardial mass proceeds slowly relative to diminution in cavity size [3]. The persistence of increased muscle mass in the setting of acutely diminished ventricular volume results in increased ventricular wall thickness and decreased cavity dimensions [4]. This abrupt change in the mass-tovolume relationship of the ventricle may result in significant alterations in both systolic and diastolic function of the ventricle. The recognition that a strikingly high wall thickness-to-cavity volume ratio is a frequent echocardiographic finding in patients exhibiting a low cardiac output state early after Fontan operation [5], and a frequent pathologic finding in postmortem specimens of nonsurvivors, has led to an appreciation of the impact of acute alterations in ventricular geometry on the outcome of Fontan operations. We hypothesized that dividing the Fontan operation into two procedures could accomplish earlier reduction 1996 by The Society, of Thoracic Surgeons Published by Elsevier Science Inc PII S3-4975(96)366-9

2 Ann Thorac Surg JACOBS ET AL ;62: HEMI-FONTAN OPERATION of the volume work of the single ventricle, and might minimize the impact of changes in ventricular geometry on outcome and survival. Thus the hemi-fontan operation [6, 7] was conceived as a first step in a two-stage process of achieving total caval pulmonary connection. Obligating superior vena caval return to pass through the lungs before returning to the functional single ventricle, the hemi-fontan operation is physiologically similar to the bidirectional Glenn anastomosis [8]. Because it is performed as an intermediate step before an anticipated completion Fontan procedure, it differs technically from the bidirectional Glenn anastomosis in ways that simplify the eventual completion Fontan operation or total caval pulmonary connection [9]. In addition, other sources of pulmonary blood flow including systemic-to-pulmonary shunts and antegrade flow to the pulmonary arteries are eliminated. This study was undertaken to review our initial experience with the hemi-fontan operation as part of a strategy of early reduction of the volume work of the single ventricle and a two-staged approach to the Fontan operation. Material and Methods Patient Population We reviewed retrospectively the records of 4 consecutive patients less than 2 years of age who underwent the hemi-fontan operation at the Children's Hospital of Philadelphia from May 1989 through August The patients were grouped according to their principal cardiac diagnosis (Fig 1) including hypoplastic left heart syndrome (266 patients, 66.5%), tricuspid atresia (29 patients, 7.3%), single left ventricle (25 patients, 6.3%), complex double-outlet right ventricle (19 patients, 4.8%), atrial isomerism-heterotaxia syndrome (19 patients, 4.8%), pulmonary atresia with intact ventricular septum (1 patients, 2.5%), and other (32 patients, 8.%). All patients were then further classified as to the morphology of the single or dominant ventricle: right ventricular or left ventricular dominance. In each instance, the hemi- Fontan operation was classified as either elective or urgent. Urgent procedures were those undertaken during an unscheduled hospital admission necessitated by deteriorating clinical status, eg, severe hypoxemia or ventricular failure. Age at operation ranged from 2 months to 24 months (Fig 2), with a mean age of 8.5 months, and with 227 patients (57%) having had operation at 5, 6, or 7 months of age. Follow-up (with respect to early and late deaths) was complete to October Thus follow-up ranged from I to 76 months, with a mean follow-up of 33 months. Surgical Technique All operations were performed by two surgeons using a single protocol for operative management. All procedures were performed through a median sternotomy incision using cardiopulmonary bypass and deep hypothermic circulatory arrest. Bypass was established with ascending aortic perfusion and drainage via a single cannula in the systemic venous atrium. All systemic-to- Tricuspid Atresia t'-.. ~ P A - H e v e s rtaxia Other Fig 1. Each of the 4 patients undergoing hemi-fontan operation at age less than 2 years was classified as to principal cardiac diagnosis. (DORV = double-outlet right ventricle; HLHS = hypoplastic left heart syndrome; LV left ventricle; PA-WS = pulmonary atresia with intact ventricular septum.) pulmonary artery shunts were occluded upon initiation of bypass. After cooling to 18 to 2 C, the aorta was clamped and a single dose of crystalloid cardioplegia was infused into the aortic root. The circulation was temporarily discontinued. The hemi-fontan procedure was performed as previously described by us [6]. An incision was made in the most superior portion of the right atrium and carried superiorly onto the medial aspect of the right superior vena cava. The confluence of the right and left branch pulmonary arteries was opened anteriorly from a point just medial to the upper lobe branch on each side. In the absence of naturally occurring or surgically created pulmonary atresia, the main pulmonary artery was transected and oversewn proximally at the level of the pulmonary valve. The right superior vena cava was anastomosed in side-to-side fashion to the right pulmonary artery. If a left superior vena cava was also present, it too was opened along its medial aspect, occluded at its cardiac end, and anastomosed in side-to-side fashion to the ipsilateral branch pulmonary artery. A single patch of cryopreserved pulmonary artery homograft was used to augment the branch pulmonary arteries anteriorly, to create a roof over the anastomoses of the vena(e) cava(e) to the ipsilateral branch pulmonary arteries, and to occlude the inflow of the right superior vena cava into the right atrium (the junction of the right superior vena cava with the right atrium having been enlarged to equal or exceed the caliber of the inferior vena cava). Resumption of bypass and removal of the cross clamp were followed by rewarming to 37 C and termination of bypass. Sixty-one concomitant procedures were performed in 59 patients, including atrial septectomy (17 patients), relief of aortic arch obstruction (13 patients), proximal main pulmonary artery-to-ascending aortic anastomosis (6 patients), repair or revision of previous repair of anomalous pulmonary venous connection (6 patients), atrioventricular valvuloplasty (6 patients), and other procedures (13 patients). Statistical Analysis Early mortality was defined as death within 3 days of hemi-fontan procedure including death after hospital

3 458 JACOBS ET AL Ann Thorac Surg HEMI-FONTAN OPERATION 1996;62: Age at Hemi - Fontan (months) Fig 2. Histogram describing the distribution of the 4 consecutive patients undergoing hemi-fontan operation by age at operation. A total of 227 patients (57%) had operation at 5, 6, or 7 months qf age. discharge. Univariate analysis of the relationships of individual preoperative variables to early death after hemi-fontan operation was performed with Fisher's exact test. This is inclusive of age as a dichotomous variable. The influence of age as a continuous variable was evaluated by Wilcoxon rank sum test. Age at operation and dichotomous variables including ventricular dominance, hypoplastic left heart syndrome (versus other diagnosis), concomitant procedures, and urgent operation were entered in a stepwise logistic regression to identify independent risk factors for mortality. A p value less than.5 was considered significant. Kaplan-Meier estimates were computed to show the actuarial survival rate. Results Among 4 patients, there were 31 deaths in the 3 days after operation for an early mortality rate of 7.8%. Early mortality (Fig 3) was significantly lower in the second half of the series (patients 21 to 4, eight early deaths, mortality 4%; p <.1) and there was one early death in the last 1 consecutive patients. In the univariate analysis, age at operation (p =.2), concomitant procedure (p -<.1), and urgent operation (p <.1) were predictors of early mortality. Hypoplastic left heart syndrome (versus other diagnosis) and ventricular dominance were not significantly related to early mortality. In the multivariate logistic regression analysis, only concomitant procedures and urgent operation were significantly related to early mortality. Age at operation was not an independent predictor of early mortality. The relationship of the age at operation to mortality was further investigated as follows: patients who had undergone urgent operation (n = 14) were deleted from the analysis, leaving 386 patients who underwent elective herni-fontan operations. Those who underwent hemi- Fontan operation at age less than 6 months (85 patients; mortality, 8.2%) were compared with those who underwent hemi-fontan operation at age 6 months or greater (31 patients; mortality, 5.7%). Mortality was not significantly different (p =.44). Similarly, those who underwent hemi-fontan operation at age less than 5 months (27 patients; mortality, 11.1%) were compared with those who underwent hemi-fontan operation at age 5 months or greater (359 patients; mortality, 5.9%). Mortality was not significantly different (p =.23). Also of note is the lower mean age of patients in the second half of the series: patients 21 to 4 had a mean age of 7.5 months versus patients 1 to 2, who had a mean age of 9.5 months. This is noteworthy in relation to the significantly lower mortality rate in the second half of the series, despite equal prevalence of concomitant procedures and urgent operations in the first and second halves of the series. Although elective hemi-fontan operation at age less than 5 months was not associated with increased mortality compared with that of patients aged 5 months or older, it was the case that urgent operation with one or more concomitant procedures at age less than 5 months was associated with a prohibitively high mortality (5 of 6 patients, mortality-83%). By way of contrast, 4 patients with no prior palliative procedures underwent hemi- Fontan operation electively at age less than 3 months with no early or late mortality. Survivors of the hemi-fontan operation were subjected to hemodynamic evaluation by cardiac catheterization 6 to 18 months after hemi-fontan operation. Only 8 patients were thought to have hemodynamics unsuitable for eventual completion of the Fontan operation. Of these, 4 were listed for heart transplantation, with 3 patients undergoing four heart transplant operations with 2 long-term survivors. One patient died after tricuspid valvuloplasty 2 months after hemi-fontan operation. One patient with elevated pulmonary vascular resistance died in an automobile accident. Two patients with chronic pulmonary insufficiency remain stable ::m I 23 n = 1 to n = 21-4 * p<.1 Fig 3. Early mortality (<3 days alter hemi-fontan operation) was separately computed for the first half (patients 1 to 2) and the second half (patients 21 to 4) of the series. The shaded columns represent patients within each cohort who died within 3 days of hemi-fontan operation. The early mortality was significantly less for the second group of patients (p <.1; see text).

4 Ann Thorac Surg JACOBS ET AL ;62: HEMI-FONTAN OPERATION r J Months Fig 4. Kaplan-Meier plot of actuarial survival rate after hemi-fontan operation. Broken lines indicate 7% confidence limits. though dependent on mechanical ventilation 2 and 3 years after hemi-fontan operation. Analysis of actuarial survival rate, shown in the Kaplan-Meier plot (Fig 4), considers all early and late postoperative deaths inclusive of early and late mortality associated with subsequent completion Fontan procedures. Comment Reduction of the volume work of the single-ventricle heart must be accomplished not only with low operative mortality, but sufficiently early in life to minimize the effects of chronic volume overload on ventricular geometry and myocardial ultrastructure. The hemi-fontan operation as described was associated with a 3-day mortality rate of 4% in the last 2 consecutive patients and 1% in the last 1 consecutive patients in the face of steadily and intentionally declining age of referral. We have hypothesized that a staged approach to Fontan operation including hemi-fontan operation early in the first year of life would favorably affect the overall survival and quality of life of patients with single-ventricle hearts by reducing the mortality associated with the condition of chronic ventricular volume overload (the unoperated or palliated state), reducing the mortality associated with the Fontan operation by achieving normalization of ventricular geometry and mass-to-volume ratio before the completion Fontan operation, and by elimination of important risk factors such as pulmonary artery distortion. That early reduction of volume overload reduces mortality associated with the unoperated or palliated condition is difficult to prove short of a randomized study, which would be difficult to justify on ethical grounds. Franklin and associates [1] reviewed the survival without definitive repair in 191 patients presenting with double-inlet ventricle before 1 year of age. Although the true natural history without intervention could not be determined because of study design, actuarial survival was 57% at 1 year and 43% at 5 years. Certainly experience with hypoplastic left heart syndrome at this institu- 48 tion and elsewhere has led to the recognition of a significant time-related risk of mortality associated with the palliated state [11, 12]. Interposition of the hemi- Fontan operation at about 6 months of age has resulted in an increased percentage of survivors of initial Norwood stage I palliation ultimately reaching physiologic repair by Fontan operation I6]. The altered relationship of ventricular mass-to-volume after an operation that reduces volume work of the systemic ventricle has been described in biventricular hearts after closure of ventricular septal defects [13] and in hearts with functional single ventricles after the Fontan operation [4], and more recently after bidirectional caval pulmonary anastomosis [14] and hemi-fontan operation [15]. The negative impact of ventricular hypertrophy on survival after Fontan operation has long been recognized, but only more recently has it been understood in relation to the diastolic dysfunction that accompanies abrupt removal of the excessive volume load associated with palliated single-ventricle circulation [16]. Two separate phenomena favor the performance of the hemi-fontan operation as an interim procedure to reduce the volume work of the ventricle before Fontan operation. One is the observation that although the volume work of the ventricle is reduced from the sum of pulmonary and systemic blood flow to systemic blood flow only after hemi-fontan operation, just as it is after Fontan operation, the low output state associated with contracted ventricular cavity volume is only rarely observed and considerably less lethal after hemi-fontan operation than it is after a primary Fontan operation [17]. This is partially explained by the fact that after hemi- Fontan operation only superior vena caval flow traverses the pulmonary vascular bed, impeding ventricular filling to a lesser extent than does the Fontan operation, wherein the complete cardiac output must passively traverse the pulmonary vasculature without the impelling force of a ventricular pump. The second phenomenon is the empiric observation described by Rychik and colleagues [15] that although ventricular geometry is acutely altered after both hemi-fontan operation and Fontan operation, the degree of change in wall thickness, cavity volume, and wall thickness-to-cavity diameter ratio is of significantly lesser magnitude after the hemi- Fontan operation. Whether the duration of exposure of the single ventricle to excessive volume work ultimately affects ventricular function late after the Fontan operation is not known. However, the observation of structural myocyte changes in experimental models of ventricular volume overload related to chronic mitral regurgitation [18], together with changes in cardiac collagen and elastin observed in experimental models of volume overload related to arteriovenous shunts [19], suggests that myocardial development is likely to be altered in infants with heart malformations characterized by a functional single ventricle. Concerning the reversibility of changes in morphology and function in relation to the duration of volume overload, observations by Cordell and associates [13] concerning the effects of corrective operation on left heart

5 46 JACOBS ET AL Ann Thorac Surg HEMI-FONTAN OPERATION 1996;62: volume and mass in children with ventricular septal defect are consistent with the hypothesis that normalization of myocardial mass and ventricular function is more likely after removal of the volume load in infancy than in later years. A potential advantage of a two-staged approach to the Fontan operation is the technical simplification of the eventual completion Fontan procedure. Conversion from hemi-fontan to total caval pulmonary connection requires only the excision of the patch occluding inflow from the superior vena cava to the right atrium together with construction of a lateral atrial tunnel. Alternatively, interposition of an extracardiac tube graft from the inferior vena cava to the homograft roof over the anastomosis of the right superior vena cava to the pulmonary arteries completes an "extracardiac" Fontan operation. In the latter case, fenestration may be achieved simply by creating a hole in the portion of the homograft patch that occludes the junction of the right atrium with the right superior vena cava [2]. Another advantage of a twostaged approach to the Fontan operation may be the elimination of some risk factors by the hemi-fontan operation. When the outcome of two-staged Fontan operation in 148 patients (a subset of the present series of 4 patients) was analyzed in relation to preoperative hemodynamic and angiographic characteristics by Apostolopoulou and associates (personal communication), multiple regression analysis identified only elevated ventricular end-diastolic pressure and significant atrioventricular valve regurgitation (before hemi-fontan operation) as risk factors for mortality before or after completion Fontan operation. Some traditional risk factors, including pulmonary artery distortion, appeared to have been "neutralized" by interposition of the hemi- Fontan operation before the Fontan operation. Our institutional experience with patients undergoing Fontan operations for heart malformations with singleventricle physiology has been characterized by steadily declining rates of operative mortality associated with the Fontan operation. Norwood and Jacobs' 1993 report [21] of the Fontan procedure in two stages revealed a decrement in hospital mortality for the Fontan operation from 16% to 8%. Additional technical modifications were associated with a further reduction of the mortality to 4.5% among 112 successive patients who underwent completion Fontan operation [9]. Most recently, 46 patients have undergone completion Fontan operation between July 1994 and October 1995 with one hospital death (2.4%) and one late death (2.4%). Although reduction in the mortality associated with the Fontan operation for functional single ventricles is certainly multifactorial, the use of a two-staged approach and early referral for elective hemi- Fontan procedure (at 5 to 8 months of age) emerge as important elements of a strategy that has been associated with a marked improvement in outcome. Additional long-term follow-up, including hemodynamic assessment and estimation of exercise capacity, will be necessary to further assess the hypothesis that early reduction of the volume work of the single ventricle is an important ingredient of a management strategy based on the Fontan operation. We gratefully acknowledge the assistance of Dr Chao Chen, Dr Gail Anolik, Ms Lynn Wang and Ms. Nancy Harnum of the Department of Surgery, Deborah Heart and Lung Center, in statistical analysis and graphic presentation of the data. References 1. Sluysmans T, Sanders SP, Van Der Velde M. Natural history and patterns of recovery of contractile function in single left ventricle after Fontan operation. Circulation 1992;86: Kirklin JK, Blackstone EH, Kirklin JW, Pacifico AD, Bargeron LM. The Fontan operation. Ventricular hypertrophy, age and date of operation as risk factors. J Thorac Cardiovasc Surg 1986;92: Papadimitrious JM, Hopkins BE, Taylor RR. Regression of left ventricular dilation and hypertrophy after removal of volume overload. Circ Res 1974;35: Chin AJ, Franklin WH, Andrews BAA, Norwood WI Jr. Changes in ventricular geometry early after Fontan operation. Ann Thorac Surg 1993;56: Farrell PE, Chang AC, Murdison KA, Baffa JM, Norwood WI, Murphy JD. Outcome and assessment after the modified Fontan procedure for hypoplastic left heart syndrome. Circulation 1992;85: Norwood WI Jr, Jacobs ML. Fontan procedure for hypoplastic left heart syndrome. Ann Thorac Surg 1992;54: Douville EC, Sade RM, Fyfe DA. Hemi-Fontan operation in surgery for single ventricle: a preliminary report. Ann Thorac Surg 1991;51: Hopkins RA, Armstrong BE, Serwer GA, Peterson RJ, Oldham NH. Physiologic rationale for a bidirectional cavopulmonary shunt. J Thorac Cardiovasc Surg 1985;9: Jacobs ML, Norwood WI Jr. Fontan operation: influence of modifications on morbidity and mortality. Ann Thorac Surg 1994;58: Franklin RCG, Spiegelhalter DJ, Anderson RH, et al. Double-inlet ventricle presenting in infancy. I. Survival without definitive repair. J Thorac Cardiovasc Surg 1991;11: Barber G, Murphy JD, Pigott JD, Norwood WI Jr. The evolving pattern of survival following palliative surgery for hypoplastic left heart syndrome [Abstract]. J Am Cardiol 1988;2:139A. 12. Jonas JA. Intermediate procedures after first-stage Norwood operation facilitate subsequent repair. Ann Thorac Surg 1991;52: Cordell D, Graham TP, Atwood GF, Boerth RC, Boucek RJ, Bender HW. Left heart volume characteristics following ventricular septal defect closure in infancy. Circulation 1976; 54: Berman NB, Kimball TR. Systemic ventricular size and performance before and after bidirectional cavopulmonary anastomosis. J Pediatr 1993;122:$ Rychik J, Jacobs ML, Norwood WI Jr. Acute changes in left ventricular geometry after volume reduction operation. Ann Thorac Surg 1995;6: Gewillig M, Daene W, Aubert A, Van der Hauwaert L. Abolishment of chronic volume overload: implications for diastolic function of the systemic ventricle immediately after Fontan repair. Circulation 992;86(Suppl 2): Lamberti JJ, Mainwaring RD, Spicer RL, Uzark KC, Moore JW. Factors influencing perioperative morbidity during palliation of the univentricular heart. Ann Thorac Surg 1995;6: $ Spinale FG, Ishihra K, Zile M, DeFryte G, Crawford FA, Carabello BA. Structural basis for changes in left ventricular function and geometry because of chronic mitral regurgita-

6 Ann Thorac Surg JACOBS ET AL ;62: HEM1-FONTAN OPERATION tion and after correction of volume overload. J Thorac Cardiovasc Surg 1993;16: Ruzicka M, Keeley FW, Leenen FHH. The renin-angiotensin system and volume overload-induced changes in cardiac collagen and elastin. Circulation 1994;9: Hennein HA, Kililtan HT, Sade RM. Conversion of the hemi-fontan procedure to fenestrated total extracardiac cavopulmonary bypass. Ann Thorac Surg 1995;6:$ Norwood WI, Jacobs ML. Fontan's procedure in two stages. Am J Surg 1993;166: DISCUSSION DR FRANCISCO J. PUGA (Rochester, MN): This is a very excellent presentation and a very considerable experience on a difficult subject. I congratulate you for your excellent results. I have one question. Did I understand correctly that you propose interruption of any other source of pulmonary blood flow at the time of performance of the hemi-fontan or bidirectional Glenn procedure? If that is the case, have you seen any patients in whom pulmonary arteriovenous fistulas have appeared? We have seen some of this in our experience, and we have also seen development of extensive systemic-to-pulmonary collaterals that may actually interfere with the performance of the Fontan procedure later on. DR JACOBS: Yes, i think that is a very appropriate question. The interval between hemi-fontan and completion Fontan operation for the majority of these patients is on the order of 12 months. We have not seen in this group of patients evidence of important intrapulmonary arteriovenous fistulas. There are a few outliers who, on the basis of social reasons or reasons of referral, are now 3 or 31/2 years out after the hemi-fontan operation and awaiting reevaluation before completion, and I think we will have to look very critically to see if there is any evidence of arteriovenous fistula formation in that group. DR CARLO MARCELLETTI (Rome, Italy): I enjoyed very much the presentation of such extensive experience. I would like to reinforce the observation that Dr Puga has made. Excluding hypoplastic left heart syndrome, for all other congenital anomalies we prefer to leave additional sources of pulmonary blood flow when performing the hemi-fontan operation. This is because we have observed development of arteriovenous fistulas in patients with heterotaxy syndrome in whom we had done a bidirectional Glenn procedure. I would also like to reiterate what you mentioned about sudden changes in geometry of the ventricle when performance of a cavopulmonary anastomosis abruptly decreases the pulmonary-to-systemic flow ratio to 1. or less, for example, after pulmonary artery banding of a univentricular heart. In such a condition we always avoid a one-stage Fontan operation and instead go through a first-stage Damus-Kaye-Stansel procedure and cavopulmonary anastomosis, followed later by total extracardiac cavopulmonary connection. DR JACOBS: Doctor Marcelletti, I would agree completely with your latter comment. And part of the strategy in staging the Fontan operation is to eliminate any of those obstructions or hemodynamic burdens that could cause persistence of myocardial hypertrophy up to the time of the Fontan procedure. I therefore would agree that if one anticipates pulmonary arteryto-aortic anastomosis, it is very good to accomplish it at the time of the hemi-fontan operation, as it was in 6 patients in this series with no mortality among that small group. With regard to your first question, I think the development of arteriovenous fistulas in the lungs in heterotaxy syndrome patients after a Kawashima operation or total cavopulmonary connection is a fascinating quirk of nature. I am not certain whether the substrate is exactly the same as with the bidirectional cavopulmonary anastomosis or hemi-fontan operation because we have seen in the heterotaxy group, as you described, very early development of arteriovenous fistulas, less than 6 months after total cavopulmonary anastomosis. I would only encourage you and the group perhaps not to look on this as an untreatable or irreversible problem because we now have experience in a small group of patients creating an intracardiac lateral tunnel or an extracardiac Marcelletti-type Fontan operation connecting the hepatic veins to the lung arteries, and at least in 2 of those patients at recatheterization we no longer see evidence of the arteriovenous fistulas in the lungs. DR IVATURY M. RAO (Abu Dhabi, United Arab Emirate): I congratulate Dr Jacobs on a wonderful experience and presentation. I just have one small doubt about this concept of hemi-fontan operation. When you have these big incisions in and near the superior vena caval-right atrial junction and put a patch at the mouth of the superior vena cava, what is the incidence of maintenance of sinus rhythm? Also, does the patch of the superior vena cava mouth interfere with the growth of the superior vena caval-right atrial junction, when you go to do the Fontan operation? DR JACOBS: I think those are two very important questions. I will answer the second first. The second question was, does the patch closure of the right atrial-superior vena caval junction interfere with growth of that opening? The incision that is made in the superior part of the atrium and carried on to the medial aspect of the right superior vena cava allows us to close the superior vena caval-right atrial junction with a patch considerably larger than the original size of the superior vena caval-right atrial junction. Doctor Van Praagh did an interesting morphologic study where he showed that in patients with a variety of heart malformations, the caliber of the opening of the superior vena cava into the heart was often as small as 6% of the caliber of the opening of the inferior vena cava into the heart. It is specifically in relation to that phenomenon that we routinely enlarge the opening of the superior vena cava into the heart. As far as the impact of this strategy on rhythm, I think that this is a very important thing to investigate. We are presently in the process of analyzing retrospectively the impact of this staged approach on rhythm, both early and late after operation. I do not have the hard data. I would say anecdotally that the overwhelming majority of patients who come back for completion Fontan operation after a prior hemi-fontan operation are in either a sinus rhythm or a rhythm that is initially interpreted as sinus rhythm but may in fact originate from another atrial pacemaker and is often a left atrial rhythm. I think what ultimate impact that is going to have remains to be determined. DR EDWARD L. BOVE (Ann Arbor, MI): We recently reviewed our experience with 1 patients having hemi-fontan procedures before age 6 months and would agree with you that young age was not a risk factor for mortality but it seemed to be one for morbidity. I wondered what comments you might make on

7 462 JACOBS ET AL Ann Thorac Surg HEMI-FONTAN OPERATION 1996;62: postoperative morbidi~ in the very young patients from your experience. We noticed that some of these patients were fairly hypoxemic for the first few days and then seemed to get better. When we looked at postoperative arrhythmias, it prompted us to change our technique, modifying the incision. It no longer crosses the cavoatrial junction but is limited to the base of the right atrium. Since then, all of the patients have been in sinus rhythm, which we think has been a benefit. Would you comment on that? DR JACOBS: 1 think the latter is very encouraging if it is associated with consistent preservation of sinus rhythm. In terms of the very young patients, we are aware of the excellent results that you presented at the American Heart Association meeting where you showed that the saturation early postoperatively was lower in the very young patients, and I think a few of them experienced caval or pulmonary artery thrombosis. We have not seen that complication, although I would echo your observation that the initial arterial oxygen tension and systemic arterial saturation is generally lower in this group. That finding tends to persist for only the first 24 or 36 hours. By the time they leave the intensive care unit, they all wind up being about 84% saturated. So I think that the pulmonary vascular resistance probably behaves somewhat differently in the very young patients early postoperatively. Bound volumes available to subscribers Bound volumes of the 1995 issues of The Annals of Thoracic Surgery are available only to subscribers from the Publisher. The cost is $99. (outside US add $25. for postage) for volumes 59 and 6. Each bound volume contains a subject and author index, and all advertising is removed. The binding is durable buckram with the name of the journal, volume number, and year stamped on the spine. Payment must accompany all orders. Contact Elsevier Science Inc, 655 Avenue of the Americas, New York, NY 11; or telephone (212) (facsimile: (212) ).