Impact of the Evolution of the Fontan Operation on Early and Late Mortality: A Single-Center Experience of 405 Patients Over 3 Decades

Transcription

1 Impact of the Evolution of the Fontan Operation on Early and Late Mortality: A Single-Center Experience of 405 Patients Over 3 Decades Hideo Ohuchi, MD, PhD, Koji Kagisaki, MD, Aya Miyazaki, MD, Masataka Kitano, MD, Satoshi Yazaki, MD, Heima Sakaguchi, MD, PhD, Hajime Ichikawa, MD, PhD, Osamu Yamada, MD, and Toshikatsu Yagihara, MD, PhD Departments of Pediatric Cardiology and Thoracic Surgery, National Cerebral and Cardiovascular Center, Osaka, Japan Background. Postoperative mortality has decreased in patients undergoing the Fontan, and the determinants of such mortality may also have changed significantly. Methods. We conducted a study intended to focus on clarifying the determinants of mortality in 405 consecutive patients who had undergone a Fontan (62 patients after an atriopulmonary connection, 105 after an intra-atrial rerouting, and 238 patients after an extracardiac rerouting) between 1979 and Results. The overall 1-year, 5-year, 10-year, and 15-year rates of survival were 87.1%, 84.6%, 83.4%, and 81.6%, respectively, and the type of procedure as well as heterotaxy syndrome, ventricular ejection fraction, and atrioventricular valve repair at the time of the Fontan were independent predictors of overall mortality (p < 0.05 for all). Heterotaxy syndrome and atrioventricular (AV) valve repair were independent predictors of early (less than 6 months postoperative) mortality and the type of procedure was an independent predictor of late (6 or more months postoperative) mortality (p < 0.05 to 0.01). In the era of intra-atrial rerouting, heterotaxy syndrome was the only independent predictor of total, early, and late mortality (p < 0.05 for all), whereas a low ejection fraction, AV valve repair, and repair of a total anomalous pulmonary vein connection, rather than heterotaxy syndrome, were independent predictors of total or early mortality or both (p < 0.05 for all) in the era of extracardiac rerouting. Conclusions. Even in the modern era of extracardiac rerouting in the Fontan, a low ventricular ejection fraction, AV valve dysfunction, or a total anomalous pulmonary vein connection remain significant risk factors for mortality in patients with a single-ventricle physiology. (Ann Thorac Surg 2011;92: ) 2011 by The Society of Thoracic Surgeons In 1992, Driscoll and colleagues reported the comprehensive clinical profiles, including the early and late mortality, of 352 patients who had undergone a Fontan with an atriopulmonary connection (APC), and demonstrated that only one third showed significant clinical improvement and that another one third died within 5 years after having the. On the basis of those results, Driscoll and colleagues emphasized the importance of the selection criteria for the [1]. However, remarkable improvements in the Fontan procedure have been introduced since the initial report of its use, such as the introduction of the total cavopulmonary connection through intra-atrial rerouting (IAR) in 1988 by de Leval and associates [2], a staged approach with a Glenn anastomosis in 1989 [3], and the version of extracardiac rerouting (ECR) introduced by Marcelletti and Accepted for publication May 9, Address correspondence to Dr Ohuchi, Department of Pediatric Cardiology, National Cerebral and Cardiovascular Center, 5-7-1, Fujishiro-dai, Suita, Osaka , Japan; hohuchi@hsp.ncvc. go.jp. coworkers in 1990 [4]. In addition to the baffle fenestration introduced in the 1990s [5], postoperative management with nitric oxide has made a significant positive contribution to the reduction in mortality among patients undergoing the Fontan procedure [6]. The current trend toward early surgical intervention with the goal of early volume reduction of the systemic ventricle (SV), and the evolution in the decade after 2000 of pharmacotherapy for the pulmonary circulation [7], may also have contributed significantly to the recent low postoperative mortality among patients undergoing the Fontan procedure. As a result, the classic recommendations for the made by Choussat and associates [8] may no longer be suitable for ECR in the current era of the Fontan procedure [9], and the clinically relevant determinants of mortality after the procedure may and seemingly should have changed over the course of time. Accordingly, we conducted a study to clarify changes in the clinical determinants of total, early, and late mortality after the Fontan on the basis of the type of Fontan procedure used. Although we were also aware of the importance of postoperative morbidity, such as arrhythmias, in patients undergoing the procedure, we focused 2011 by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 1458 OHUCHI ET AL Ann Thorac Surg EVOLUTION OF THE FONTAN OPERATION 2011;92: on the mortality-related issues in the study partly because of limitations in time and the scope of reporting. Patients and Methods Subjects We retrospectively reviewed a total of 405 consecutive patients (245 males and 160 females) with congenital heart disease who had undergone a Fontan between October 1979 and May 2010 (Table 1), and for whom the last follow-up visit took place at the end of December The morphology of the patients SVs was categorized into three groups based on the dominant ventricle, as previously described [10]. A left ventricular (LV) type SV was present in 159 patients, a biventricular type SV in 72, and a right ventricular (RV) type SV in 174. The major cardiac diagnoses were a univentricular heart in 109 patients, followed by tricuspid atresia, double Table 1. Patient Characteristics Total outlet right ventricle, mitral atresia, atrioventricular (AV) septal defect, pulmonary valve atresia, hypoplastic left heart, and other, less frequent diagnoses. Among all 405 patients in the study, heterotaxy syndrome (HS) and a common AV valve were seen in 136 (33.6%) and 119 patients (29.4%), respectively. We observed a significant transition in the type of Fontan, from an APC- to an ECR- through an IAR-type, in the patients enrolled in our study according to the time at which the procedure was done, and our current procedure is an ECR-type Fontan. Of the patients with an ECR-type procedure, 126 (52.9%) underwent the ECR in an off-pump fashion. The number of surgical procedures per patient, including palliative ones, ranged from 1 to 7, and a primary Fontan was performed in 62 patients. Among the patients palliative pre-fontan procedures, the major procedure was the creation of an aortopulmonary shunt, Era of First Fontan Operation 1980s 1990s 2000s ( ) ( ) ( ) ( ) n Age at Fontan (years) SV type (LV/BV/RV) 159/72/174 37/8/22 56/24/84 66/40/68 Diagnosis Heterotaxy (30%) 69 (42%) 47 (27%) UVH TA DORV MA CAVC PA HLHS Others Type of repair APC IAR ECR Previous or additional procedures at time of Fontan APS PAB Glenn AVVP DKS TAPVC repair Coil embolization for collaterals Fenestration Values are mean SD. APC atriopulmonary connection; APS aortopulmonary shunt; AVVP atrioventricular valvuloplasty; BV biventricle; CAVC common atrioventricular canal; DORV double-outlet right ventricle; DKS Damus-Kay-Stansel procedure; ECR extracardiac rerouting; HLHS hypolastic left heart syndrome; IAR intra-atrial rerouting; LV left ventricle; MA mitral valve atresia; PA pulmonary valve atresia; PAB pulmonary artery banding; RV right ventricle; SV single ventricle; TA tricuspid valve atresia; TAPVC total anomalous pulmonary vein connection; UVH univentricular heart.

3 Ann Thorac Surg OHUCHI ET AL 2011;92: EVOLUTION OF THE FONTAN OPERATION 1459 such as a Blalock-Taussig (BT) shunt, in 55.6% of the patients, followed by a Glenn anastomosis in 49.6%, coil embolization for minimizing aortopulmonary collateral flow in 36.0%, pulmonary artery banding, and atrioventricular valve repair (AVVR), including the use of a mechanical AV valve replacement, in the remaining patients. This study protocol was approved by the Ethics Committee of the National Cerebral and Cardiovascular Center. Definition of Early and Late Postoperative Mortality Because the postoperative mortality was high at 6 months after the [1, 11], and it might not have been very difficult to prolong patients lives with currently used advanced intensive care procedures through a device system that might have obscured early mortality, we defined early mortality as death at less than 6 months postoperatively and late mortality as death beyond 6 months postoperatively. Clinical Variables The clinical variables included the era in which the patient s Fontan was done (1980s, 1990s, and 2000s), as well as the type of Fontan (APC, IAR, and ECR), age of the patient at the time of the first Fontan, SV morphology (LV, biventricular, and RV), HS, preoperative SV function, mean pulmonary artery pressure, fenestration at the time of the Fontan, Glenn anastomosis done before the Fontan (staged strategy), AVVR at the time of the Fontan, and repair of a total anomalous pulmonary vein connection (TAPVC) before or at the time of the Fontan. Systemic Ventricular Function and Hemodynamics Cardiac catheterization in patients under 12 years of age had been performed years before the under light sedation with pethidine and promethazine hydrochlorides. Patients 12 years of age or older were sedated with oral pentobarbital calcium. With regard to SV morphology, the groups categorized as having a morphology of the LV or non-lv type were used to determine the major factors associated with the cardiac ejection fraction (EF). The measures of SV function used in the study included the EF, end-diastolic volume divided by the body surface area (EVDI), and end-diastolic pressure (EDP). The SV volumes were estimated with Simpson s rule on the basis of cine ventriculography, as we previously described [10]. We excluded the size of the pulmonary artery as a variable because the current strategy for the staged Glenn anastomosis made it difficult to accurately measure the diameter of the anastomosed pulmonary artery. It was also difficult to accurately estimate the resistance of the pulmonary artery because of the absence of a ventricle that mixes the systemic venous return as well as the possible existence of venovenous collateral vessels [12]. Our traditional policy of treating AV valvular regurgitation was based on the grade of regurgitation. If the patient had moderate or greater AV valvular regurgitation, repair was performed at the time of the Fontan. The severity of AV valvular regurgitation was estimated by Doppler color flow mapping and was graded as none to mild, moderate, and severe. A grade of mild implied that the AV valvular regurgitation jet crossed less than two-thirds of the systemic atrium; a grade of moderate implied that it reached beyond twothirds of the systemic atrium; and a grade of severe implied that the jet reached the posterior wall of the atrium at a significant width. Statistical Analysis The study data are expressed as mean standard deviation. We used a univariate Cox proportional hazards model to predict the associations of the clinical factors with mortality (ie, the era in which the Fontan was performed), the patient s age at the time of the first Fontan, the type of Fontan performed (APC, IAR, ECR), fenestration, HS, type of SV morphology (RV, biventricular, LV), cardiac function and hemodynamics (EF, EDVI, EDP, and pulmonary artery pressure), and previous or additional procedures or both (AVVR, TAPVC repair, or Glenn procedure before the Fontan ). In the patients with an ECR-type, an off-pump procedure was also included in the clinical factors in the proportional hazards model. The variables that proved to be significant predictors of outcome in the univariate analysis (p 0.05) were included in the multivariate analysis done with the Cox regression model to determine any independent predictors of overall and late mortality with a hazard ratio (HR) and 95% confidence interval (CI). A multivariate logistic regression model was used to identify the independent predictors of early mortality with an OR and 95% CI. The status of a variable as being independent of the overall and late mortality was estimated with the Kaplan-Meier method, and the differences in the study groups event-free status were assessed with log-rank tests. A value of p 0.05 was considered statistically significant (Stat-View 5.0; Abacus Concepts, Berkeley, CA). Results The clinical profiles based on the era in which the Fontan was done are summarized in Table 1. Era and Fontan Operation The age of patients at the time of the Fontan decreased steadily despite a lack of change in the complexity of their disease, such as in the percentage of patients with a univentricular heart, HS, or both. Type of Fontan Operation and Additional Procedures As in many other institutions, our current strategy for the Fontan consists of performing the final ECR after a staged Glenn anastomosis followed by a coil embolization to diminish any aortopulmonary collateral flow. Some patients required a Damus-Kay-Stansel procedure for preventing future subaortic stenosis. Although

4 Table 2. Univariate and Multivariate Predictors of Overall, Early, and Late Mortality in all Patients After the Fontan Operation All (n 405) Total Mortality All (n 405) Early Mortality ( 30 days) Early Mortality All (n 360) Late Mortality Variables HR 95% CI p Value OR 95% CI p Value OR 95% CI p Value HR 95% CI p Value Univariate analysis Era of Fontan vs 2000s 1980s s Type of repair vs ECR APC IAR Age at time of first Fontan Fenestration Heterotaxy syndrome Type of SV vs LV BV RV EF EDVI EDP PAP AVVR at time of Fontan TAPVC repair Staged Glenn procedure Multivariate analysis (n 388) (n 388) (n 389) (n 360) Type of repair vs ECR APC IAR Heterotaxy syndrome EF AVVR at Fontan APC atriopulmonary connection; AVVR atrioventricular valve repair; BV bivalvular; CI confidence interval; ECR extracardiac rerouting; EDP end-diastolic pressure; EDVI end-diastolic volume index; EF ejection fraction; HR hazard ratio; IAR intra-atrial rerouting; LV left ventricle; OR odds ratio; PAP mean pulmonary artery pressure; RV right ventricle; SV single ventricle; TAPVC total anomalous pulmonary vein connection OHUCHI ET AL Ann Thorac Surg EVOLUTION OF THE FONTAN OPERATION 2011;92:

5 Ann Thorac Surg OHUCHI ET AL 2011;92: EVOLUTION OF THE FONTAN OPERATION 1461 the prevalence of a TAPVC repair has remained constant, that of AVVR has slightly increased. Mortality and Mode of Death During a mean follow up of years (median, 9.1years, range, 0 to 31.3 years), 74 patients (18.3%) died. Of these, 45 patients died at less than 6 months after the Fontan, and 23 of these 45 early deaths occurred at less than 30 days after the. Twenty-nine patients died at 6 months or more after the. The reasons for the deaths in both groups were a postoperative low cardiac output syndrome in 30 patients, heart failure in 18 patients, arrhythmias in 10 patients, hematologic events such as bleeding in 6 patients, infection and sudden death in 3 patients each, a protein-losing enteropathy in 2 patients, and an unknown cause in 2 patients. The reasons for the 29 late deaths were heart failure in 14 patients, arrhythmias in 5 patients, hematologic events and sudden death in 3 patients each, infection in 2 patients, and a protein-losing enteropathy and unknown cause of death in 1 patient each. Predictors of Overall Mortality in All Fontan Groups Table 2 summarizes the predictors of overall, early, and late mortality in all of the patients who had Fontan procedures. In the case of overall mortality, the era of the, type of procedure (Fig 1a), HS (Fig 1b), SV function, and additional surgical procedures, such as AVVR (Fig 1c), were associated with mortality. Of these factors, the type of procedure, HS, EF, and AVVR were independent predictors of mortality. Pulmonary artery pressure was not a predictor of mortality. In the case of early mortality, the univariate logistic regression model revealed that the era and patient s age at the time of the first Fontan, type of procedure, HS, SV function, AVVR, and staged Glenn anastomosis were associated with mortality. According to the multivariate logistic regression model, HS and AVVR were independent predictors of mortality. In the case of late mortality, the era and type of Fontan (Fig 1d) predicted mortality, and none of these variables were the independent predictors of mortality. Tables 3, 4, and 5 summarize the predictors of overall, early, and late mortality in the patients who had APC-type, IAR-type, and ECRtype Fontan s, respectively. Predictors of Overall Mortality in Patients With APC Operations In patients who had an APC-type Fontan, HS, EDVI, and AVVR were the predictors of overall mortality. Fig 1. (a) The Kaplan-Meier survival-rate curves for all types of Fontan s (ie, atriopulmonary connection [APC], intra-atrial rerouting [IAR], and extracardiac rerouting [ECR]. (b) Curves for patients with and without heterotaxy syndrome. (c) Curves for patients who required atrioventricular valve repair (AVVR) at the time of the Fontan. (d) Curves for late survivors (6 or more months after Fontan ) in patients with atriopulmonary connection, intra-atrial rerouting, and extracardiac rerouting. The vertical bars represent 95% confidence intervals.

6 1462 OHUCHI ET AL Ann Thorac Surg EVOLUTION OF THE FONTAN OPERATION 2011;92: Table 3. Univariate and Multivariate Predictors of the Overall, Early, and Late Mortality After APC-Type Fontan Operation APC (n 62) Total Mortality Early Mortality Late Mortality Variables HR 95% CI p Value OR 95% CI p Value HR 95% CI p Value Univariate Analysis Age at time of first Fontan Heterotaxy syndrome Type of SV vs LV BV RV EF EDVI EDP PAP AVVR at Fontan Staged Glenn procedure Multivariate analysis Heterotaxy syndrome EDVI AVVR before/at time of Fontan APC atriopulmonary connection; AVVR atrioventricular valve repair; BV bivalvular; CI confidence interval; EDP end-diastolic pressure; EDVI end-diastolic volume index; EF ejection fraction; HR hazard ratio; LV left ventricle; OR odds ratio; PAP mean pulmonary artery pressure; RV right ventricle; SV single ventricle. Of these, EDVI and AVVR were independent predictors of mortality. Heterotaxy syndrome and EDVI were the only predictors of early and late mortality, respectively. Predictors of Overall Mortality in IAR Patients In patients who had an IAR-type Fontan, HS, AVVR, and TAPVC repair were the predictors of overall mortality. Of these, HS was an independent predictor of mortality, and was the only independent predictor of both early and late mortality. Predictors of Overall Mortality in ECR Patients In patients who had an ECR-type Fontan, as for the total mortality, EF and TAPVC repair were independent predictors of overall mortality, and HS was not associated with mortality. Both EF and AVVR were independent predictors of early mortality. However, no variables, including off-pump procedures, were associated with late mortality in this group. Impacts of the TAPVC repair, AVVR at the time of Fontan, and low EF ( 0.58) on the total mortality are shown in Figures 2a c. Impact of HS on Type of Fontan Operation An adverse effect of HS on each type of Fontan procedure is shown in Figures 3a c, and this effect decreased dramatically over the successive eras of the. Determinants of EF Before ECR We attempted to clarify the factors determining the preoperative EF because this variable was relevant to the overall and early mortality in the group that had ECRtype Fontan s. We found that EF was inversely correlated with age at the time of the first Fontan (r 0.213, p 0.001). Also associated with a low EF were HS (HS: EF 54 11%, non-hs: EF , p ), a non-lv SV morphology (LV: EF , non-lv: EF 55 10%, p ), and a fenestration (fenestration: EF 54 11%, non-fenestration: EF , p ). Of these, the multivariate regression model reviealed that a late Fontan (p 0.001), HS (p 0.013) and a non-lv SV morphology (p ) were independently associated with a low EF. Comment The present study clearly confirmed that the latest modification of the Fontan, the ECR-type Fontan procedure, reduced overall, early, and late postoperative mortality in patients who had undergone the, and that the survival of patients with HS had improved considerably from the era of the first Fontan procedure, with HS ceasing to be an independent risk factor for mortality in the era of the ECR-type Fontan procedure. Instead, impaired SV function as well as AVVR and TAPVC repair are emerging as

7 Ann Thorac Surg OHUCHI ET AL 2011;92: EVOLUTION OF THE FONTAN OPERATION 1463 Table 4. Univariate and Multivariate Predictors of Overall, Early, and Late Mortality After IAR-Type Fontan Operation IAR (n 105) Total Mortality Early Mortality Late Mortality Variables HR 95% CI p Value OR 95% CI p Value HR 95% CI p Value Univariate analysis Age at time of first Fontan Fenestration Heterotaxy syndrome Type of SV vs LV BV RV EF EDVI EDP PAP AVVR at time of Fontan TAPVC repair Staged Glenn procedure Multivariate analysis Heterotaxy syndrome AVVR atrioventricular valve repair; BV bivalvular; CI confidence interval; EDP end-diastolic pressure; EDVI end-diastolic volume index; EF ejection fraction; HR hazard radio; IAR intra-atrial rerouting; LV left ventricle; OR odds ratio; PAP mean pulmonary artery pressure; RV right ventricle; SV single ventricle; TAPVC total anomalous pulmonary vein connection. important risk factors for mortality in patients undergoing the, and that more effort is required toward reducing their effect and producing a better outcome in patients with an SV physiology who have these risk factors. With further regard to the determinants of SV function, a non-lv type SV morphology and the presence of HS may be becoming more significant risk factors in the era of the ECR-type Fontan procedure with a strategy of early SV volume load reduction, although in our study these variables were not independent risk factors for mortality. Reducing Early and Late Mortality Every leading institute has reported recently improved survival after the Fontan (13 15). In the era of the APC-type Fontan procedure, the 1-year, 5-year, 10- year, and 15-year survival rates were 66%, 62%, 58%, and 56%, respectively, which much lower than the rates in the initial report of the ideal patients for the Fontan procedure [9], indicating inappropriate selection of the, inadequate experience with it, or both. However, as with many other institutions, we found that modifications of the Fontan procedure with IAR and ECR dramatically improved patient survival (Fig 1A). With these modifications, the 1-year, 5-year, 10-year, and 15-year survival rates rose to 94%, 92%, 92%, and 92%, respectively, which were equivalent to of the rates reported by Giannico and colleagues and Kim and associates [16, 17]. Surprisingly, no deaths were reported among patients undergoing the ECR-type procedure in d Udekem s series, although the number of patients was relatively small [15]. The recent evolution of modifications of the Fontan procedure has definitely improved the postoperative survival of patients with an SV physiology. However, the evidence for an excellent outcome with the IAR-type modification implies that other important risk factors for mortality exist among patients undergoing the Fontan [18]. Risk Factors for Early Mortality Other Than Type of Repair The presence of HS has been widely recognized as an important postoperative risk factor in patients undergoing the Fontan [1, 14, 19, 20]. Among the main reasons for the high mortality rate in patients with HS may be the complexity of the combined cardiac structural anomalies in such patients, such as an RV-type SV morphology, a common AV valve, TAPVC, and an intrinsic susceptibility to arrhythmias [21], all of which have been considered as risk factors doe mortality. In our 136 patients with HS, a non-lv type SV morphology, common AV valve, and TAPVC were present in 119 patients (87.5%), 104 patients (76.5%), and 26 patients (19.1%), respectively, and these prevalence rates were significantly greater than those in patients without HS, in whom the corresponding percentages were 52.8%, 5.6%, and 0%, respectively. However, HS itself was not a significant risk factor for mortality; instead, specific le-

8 1464 OHUCHI ET AL Ann Thorac Surg EVOLUTION OF THE FONTAN OPERATION 2011;92: Table 5. Univariate and Multivariate Predictors of Overall, Early, and Late Mortality After ECR-Type Fontan Operation ECR (n 238) Total Mortality Early Mortality Late Mortality Variables HR 95% CI p Value OR 95% CI p Value HR 95% CI p Value Univariate Analysis Age at time of first Fontan , Off-pump Fenestration Heterotaxy Type of SV vs LV BV RV EF EDVI EDP PAP AVVR repair at Fontan TAPVC repair Staged Glenn procedure Multivariate Analysis EF AVVR at time of Fontan TAPVC repair AVVR atrioventricular valve repair; BV bivalvular; CI confidence interval; ECR extracardiac rerouting; EDP end-diastolic pressure; EDVI end-diastolic volume index; EF ejection fraction; HR hazard radio; LV left ventricle; OR odds ratio; PAP mean pulmonary artery pressure; RV right ventricle; SV single ventricle; TAPVC total anomalous pulmonary vein connection. sions, such as an AVVR or TAPVC repair, emerged as important risk factors from within the ambiguous pathophysiology of HS. Interestingly, a history of AVVR before the Fontan had no significant adverse effect on mortality, implying the importance of better control of the AV valvular regurgitation before the Fontan in controlling the mortality associated with it. In contrast, patients with an SV physiology who required a Fig 2. Effect of (a) repair of a total anomalous pulmonary vein connection (TAPVC) and (b) atrioventricular valve repair (AVVR) at the time of the Fontan, and (c) of a low ejection fraction (EF) of the systemic ventricle inpatients with a Fontan with. The patients with a Fontan done with extracardiac rerouting (ECR) (c) were divided into two groups according to the median value (0.57) of their ejection fraction. The vertical bars represent 95% confidence intervals.

9 Ann Thorac Surg OHUCHI ET AL 2011;92: EVOLUTION OF THE FONTAN OPERATION 1465 Fig 3. Effect of heterotaxy syndrome on the overall survival rate in patients with Fontan s with (a) atriopulmonary connection (APC), (b) intra-atrial rerouting (IAR), and (c) extracardiac rerouting (ECR). The vertical bars represent 95% confidence intervals. TAPVC repair remained as challenges in completing the Fontan circulation even in the era of ECR [22]. In the present study, pulmonary artery pressure was not a risk factor for mortality, which may contradict the results of previous reports [13, 17, 20]. However, as O Brien and colleagues mentioned, our recent exclusion from Fontan surgery of patients with a high pulmonary artery pressure may be partly responsible for our results [23]. In addition, the possible influence on pulmonary artery pressure of developed venovenous collateral vessels after the staged Glenn strategy [12], as well as of aortopulmonary collaterals, may obscure an association between pulmonary artery pressure and mortality. Recent advances in pharmacologic agents for the pulmonary circulation [6] may also be responsible for the lack of an association between pulmonary artery pressure and mortality. Impaired SV function, as estimated by a low EF, emerged as an important risk factor for overall and early mortality in the era of the ECR-type Fontan procedure, as it had been in past versions of the procedure [6]. Some reports have also specified a high EDP as one of the risk factors for mortality in the Fontan [1, 14]. The EDP is easily influenced by hemodynamic conditions, such as in the case of its elevation by a high preload on the SV in the situation of left-to-right shunting in complex congenital heart disease. The data from the Mayo Clinic study may support the concept of a greater cardiothoracic ratio with a high EDP (greater preload on the SV before the Fontan ) adversely affecting the outcome of the [14]. Our independent predictors of EDVI may be consistent with the results in the era of the APC-type without a staged strategy. In contrast, the assessment of SV function with less volume overloading may be crucial to predicting mortality after an ECR-type Fontan given that the state of SV loading also influences the EF. In this respect, the fact that a low EF, a non-lv type SV, and HS all independently predicted a low EF implies an intrinsic SV functional abnormality in patients with HS, as well as the importance of SV morphology even in the era of the ECR-type Fontan. Our study has several limitations. First, it was of a retrospective nature, preventing us from demonstraing a cause-effect relationship between the factors we investigated and the risk of mortality in patients having the Fontan. Second, the relatively small number of patients with APC- and IAR-type Fontan procedures was a limitation in detecting independent predictors of mortality. Third, the estimation of SV function with cine ventriculography had certain methodologic disadvantages relative to more recently developed imaging modalities such as magnetic resonance imaging. However, these newer techniques were not available in the early history of the Fontan procedure. A future study with new imaging modalities will be needed, and analyses of early and late complications, such as arrhythmias, are also needed, to assess the long-term quality of life in patients undergoing the Fontan. Conclusions Mortality after the Fontan, and especially early postoperative mortality, has decreased dramatically from what it had been in earlier eras of the, and in accord with this the risk factors for mortality have changed over the past three decades with the modifications that have been made in the Fontan procedure. However, even in the modern era of the ECR-type Fontan procedure, a low EF, AV valvular dysfunction, and TAPVC repair remain as significant risk factors for mortality. Therefore, meticulous strategies directed at preserving the function of the SV and AV valve may be crucial to further improving the outcome of patients undergoing the Fontan in the modern era of the ECR-type procedure.

10 1466 OHUCHI ET AL Ann Thorac Surg EVOLUTION OF THE FONTAN OPERATION 2011;92: References 1. Driscoll DJ, Offord KP, Feldt RH, Schaff HV, Puga FJ, Danielson GK. Five- to fifteen-year follow-up after Fontan. Circulation 1992;85: de Leval MR, Kilner P, Gewillig M, Bull C. Total cavopulmonary connection: a logical alternative to atriopulmonary connection for complex Fontan s. Experimental studies and early clinical experience. J Thorac Cardiovasc Surg 1988;96: Mazzera E, Corno A, Picardo S, et al. Bidirectional cavopulmonary shunts: clinical applications as staged or definitive palliation. Ann Thorac Surg 1989;47: Marcelletti C, Corno A, Giannico S, Marino B. Inferior vena cava-pulmonary artery extracardiac conduit. A new form of right heart bypass. J Thorac Cardiovasc Surg 1990;100: Bridges ND, Mayer JE Jr, Lock JE, et al. Effect of baffle fenestration on outcome of the modified Fontan. Circulation 1992;86: Yahagi N, Kumon K, Tanigami H, et al. Inhaled nitric oxide for the postoperative management of Fontan-type s. Ann Thorac Surg 1994;57: Uhm JY, Jhang WK, Park JJ, Seo DM, Yun SC, Yun TJ. Postoperative use of oral sildenafil in pediatric patients with congenital heart disease. Pediatr Cardiol 2010;31: Choussat A, Fontan F, Cesse P, Vallot F, Chauve A, Bricaud H. Selection criteria for Fontan s procedure. In: Anderson RH, Shinebourne EA, eds. Paediatric cardiology. Edinburgh: Churchill Livingstone, 1978: Hosein RB, Clarke AJ, McGuirk SP, et al. Factors influencing early and late outcome following the Fontan procedure in the current era. The Two Commandments? Eur J Cardiothorac Surg. 2007;31: Ohuchi H, Yasuda K, Hasegawa S, et al. Influence of ventricular morphology on aerobic exercise capacity in patients after the Fontan. J Am Coll Cardiol 2001;37: Fontan F, Kirklin JW, Fernandez G, et al. Outcome after a perfect Fontan. Circulation 1990;81: Magee AG, McCrindle BW, Mawson J, Benson LN, Williams WG, Freedom RM. Systemic venous collateral development after the bidirectional cavopulmonary anastomosis. Prevalence and predictors. J Am Coll Cardiol 1998;32: Cetta F, Feldt RH, O Leary PW, et al. Improved early morbidity and mortality after Fontan : the Mayo Clinic experience, 1987 to J Am Coll Cardiol 1996;28: Bartz PJ, Driscoll DJ, Dearani JA, et al. Early and late results of the modified fontan for heterotaxy syndrome 30 years of experience in 142 patients. J Am Coll Cardiol 2006;48: d Udekem Y, Iyengar AJ, Cochrane AD, et al. The Fontan procedure: contemporary techniques have improved longterm outcomes. Circulation 2007;116(Suppl):I Giannico S, Hammad F, Amodeo A, et al. Clinical outcome of 193 extracardiac Fontan patients: the first 15 years. J Am Coll Cardiol 2006;47: Kim SJ, Kim WH, Lim HG, Lee JY. Outcome of 200 patients after an extracardiac Fontan procedure. J Thorac Cardiovasc Surg 2008;136: Brown JW, Ruzmetov M, Deschner BW, Rodefeld MD, Turrentine MW. Lateral tunnel Fontan in the current era: is it still a good option? Ann Thorac Surg 2010;89: Kim SJ, Kim WH, Lim HG, Lee CH, Lee JY. Improving results of the Fontan procedure in patients with heterotaxy syndrome. Ann Thorac Surg 2006;82: Gentles TL, Mayer JE Jr, Gauvreau K, et al. Fontan in five hundred consecutive patients: factors influencing early and late outcome. J Thorac Cardiovasc Surg 1997;114: Miyazaki A, Sakaguchi H, Ohuchi H, et al. The clinical course and incidence of supraventricular tachyarrhythmias after extra-cardiac conduit fontan procedures in relation to an atrial situs. Circ J 2011;75: Lodge AJ, Rychik J, Nicolson SC, Ittenbach RF, Spray TL, Gaynor JW. Improving outcomes in functional single ventricle and total anomalous pulmonary venous connection. Ann Thorac Surg 2004;78: O Brien JE Jr, Marshall JA, Young AR, Handley KM, Lofland GK. The nonfenestrated extracardiac Fontan procedure: a cohort of 145 patients. Ann Thorac Surg 2010;89: INVITED COMMENTARY The ultimate paradigm for single-ventricle palliation eludes our field, yet significant strides in understanding physiology and outcomes related to our interventions are evident. The panacea for those with single-ventricle physiology will encompass stratifying patients to a particular surgical strategy at each stage based on anatomic heterogeneity, physiologic anomalies, comorbid conditions, among other factors. This approach will be paramount to obtain the best outcomes from a quality of life, neurologic, and physiologic standpoint. Ohuchi and colleagues [1] present a 3-decade experience and evolution of the Fontan. The investigators evaluated risk factors associated with early and late mortality in light of the type of Fontan (atriopulmonary, intracardiac rerouting, and extracardiac rerouting). The authors noted improved outcomes with time; however despite advances in surgical technique and perioperative care, low single-ventricle ejection fraction, atrioventricular valve dysfunction, and total anomalous pulmonary venous connection repair are associated with an increased risk of mortality. These results, coupled with the many reports in the literature over the past decade, continue to fuel the debate over the approach to single-ventricle palliation. Piecing together the puzzle is challenging; to ultimately understand the outcomes of the Fontan, the literature must be reviewed to points well before the Fontan is even accomplished. One must note the variability in surgical intervention, perioperative care, preparation of stages, including the approach to aortopulmonary collaterals, let alone the heterogeneity in the anatomic substrate itself. Despite 40 years of attempting to understand the Fontan and the approach to single-ventricle physiology, many questions remain. The ever-increasing population of patients surviving single-ventricle palliation into adulthood begets further consternation. Studies such as these are vital pieces to the puzzle. In order to attempt to account for all variables leading to a successful outcome after the Fontan, queries must start with the perinatal period, although obviously this is not practical for most studies. With this type of global thinking, failure of the Fontan might have been 2011 by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur