- Email: [email protected]
Uncontrolled Antegrade Pulmonary Blood Flow and Delayed Fontan Completion After the Bidirectional Glenn Procedure: Real-World Outcomes in China
Uncontrolled Antegrade Pulmonary Blood Flow and Delayed Fontan Completion After the Bidirectional Glenn Procedure: Real-World Outcomes in China Tao Zhang, MD, Yisheng Shi, MD, Kaihong Wu, MD, Zhongdong Hua, MD, Shoujun Li, MD, Shengshou Hu, MD, and Hao Zhang, MD, PhD Center for Pediatric Cardiac Surgery, National Center for Cardiovascular Diseases and Fuwai Hospital, Beijing, China; Department of Cardiac Surgery, Provincial Hospital Affiliated to Shandong University, Jinan, Shandong, China; Department of Cardio-Thoracic Surgery, Shouguang People’s Hospital, Shandong, China; Department of Echocardiology, National Center for Cardiovascular Diseases and Fuwai Hospital, Beijing, China; Department Cardiac Surgery, Nanjing Children Hospital, Nanjing, China
Background. Given the low rate of Fontan completion, an aggressive policy for maintaining antegrade pulmonary blood flow (AnPBF) during the bidirectional Glenn procedure (BDG) was developed for the functional single ventricle. Methods. From 2008 to 2013, 294 patients who underwent the BDG were divided into two groups: group 1 (uncontrolled AnPBF, n [ 270) and group 2 (controlled AnPBF, n [ 24). Pulmonary artery banding was performed because of the high central venous pressure in group 2. In group 1, the patients who underwent BDG from 2008 to 2012 were further divided into group DF (delayed Fontan completion, n [ 109) and group FC (Fontan completion, n [ 42). Results. The Fontan completion rate was 16.3%, and the average interval time was 2.2 ± 1.1 years. The delay of Fontan completion did not reduce body weight gain or the survival rate. Furthermore, oxygen saturation was slightly reduced in group DF. Although no impairments
of heart function were observed, the uncontrolled AnPBF in group DF resulted in an increase in ventricular enddiastolic diameter and aggravation of atrioventricular valve regurgitation over 24 months after BDG. Logistic regression analysis revealed that systemic right ventricular morphology was a risk factor for the aggravation of valve regurgitation. Conclusions. The low Fontan achievement rate is a critical issue in China. Although the patients with delayed Fontan completion exhibited an acceptable survival rate and acceptable body weight gain, uncontrolled AnPBF was associated with ventricular enlargement and aggravation of valve regurgitation. Strategies for improving the Fontan completion rate in China should be explored and could benefit outcomes.
T
the superior vena cava (SVC) pressure and result in the persistent pleural effusions and functional SV volume overload [10, 11]. The proportion of patients who eventually complete TCPC in some Asian countries is lower than that in developed Western countries. A group from India reported on 170 patients who underwent BDG operations, and only 15% of these patients received subsequent TCPC after an average interval of 48 months [12]. Most recently, a heart center located in a relatively underdeveloped area in Northern China reported the results of 68 BDG cases, and the rate of TCPC achievement was as low as 5.8% (4/68) after an interval of 3 to 5 years [13]. Low economic status and undeveloped insurance systems have contributed to the dismal Fontan completion rates. Hence, AnPBF has been maintained in many heart centers in China and India to increase and maintain arterial oxygen saturation after BDG [14]. This strategy may provide more time for the patient’s family to realize the importance of subsequent treatment and to raise the funds necessary for TCPC. Therefore, we apply an aggressive policy of maintaining AnPBF without any systemic pulmonary shunts. Here, we performed a
he bidirectional Glenn procedure (BDG) has become an effective palliative procedure and a standard intermediate step toward total cavopulmonary connection (TCPC) for patients with a functional single ventricle (SV) [1, 2]. The main sources of additional pulmonary blood flow (PBF) after BDG include systemic-to-pulmonary shunts and antegrade PBF (AnPBF). Systemic-topulmonary shunts may increase perioperative mortality, negatively affect left pulmonary artery growth, and potentially increase pulmonary vascular resistance [3, 4]. Therefore, the placement of systemic-to-pulmonary shunts is rarely advocated for patients with hypoxia after BDG. Controversy still exists regarding maintenance of AnPBF after BDG [5, 6]. Pulsatile AnPBF may improve arterial oxygen saturation, stimulate pulmonary artery growth, and prevent the formation of arteriovenous fistulas [7–9]. However, pulsatile AnPBF may also increase Accepted for publication Oct 26, 2015. Address correspondence to Dr Hao Zhang, Center for Pediatric Cardiac Surgery, Fuwai Hospital, Chinese Academy of Medical Sciences, Beijing, 100037, China; email: [email protected].
Ó 2016 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2016;-:-–-) Ó 2016 by The Society of Thoracic Surgeons
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.10.071
2
ZHANG ET AL OUTCOMES OF BDG IN CHINA
Ann Thorac Surg 2016;-:-–-
retrospective analysis of the patients with functional SV who were treated at our center. In this study, we analyzed the midterm clinical outcomes of patients with and without Fontan completion. We sought to evaluate the outcomes of this aggressive policy of maintaining AnPBF, to address the potential beneficial effects of controlled AnPBF for patients waiting for TCPC, and to emphasize the influence of uncontrolled AnPBF on patients with delayed TCPC.
Material and Methods Study Population The ethics committee of Fuwai Hospital in Beijing approved this study. We retrospectively reviewed the clinical records of 420 patients without extracardiac anomalies or chromosomal abnormalities who underwent BDG from January 2008 to December 2013. Fortyseven patients who underwent one and one-half ventricle repair, 5 patients who received modified Blalock-Taussig shunt placement after BDG, and 74 patients with pulmonary atresia were excluded. The remaining patients with SV physiology were divided into two groups: group 1 included the patients with uncontrolled AnPBF (n ¼ 270), and group 2 included the patients in whom AnPBF was controlled by pulmonary artery banding (PAB) as a result of high SVC pressure (or central venous pressure, CVP) (n ¼ 24). The patients’ characteristics are summarized in Table 1. There were no significant differences in age, sex, body weight, body surface area, or the Nakata index between groups 1 and 2. Table 1. Patient Characteristics and Anatomic Lesions at the Time of the Bidirectional Glenn Procedure
Age (years) Sex (/F) Body weight (kg) BSA (m2) PAI (mm2/m2) Single ventricle Left ventricle morphology Right ventricle morphology Biventricular or uncertain DORV TA TGA Ebstein anomaly CAVC Others
Group 1 n ¼ 270
Group 2 n ¼ 24
p value
2 (0.3–25) 173/97 13.6 9.3 0.6 0.3 252.3 125.4 64 (23.8%) 30 (11.1%)
2.1 (0.5–11) 15/9 14.7 7.9 0.6 0.2 225.5 86.5 9 (37.5%) 4 (16.7%)
0.563 0.822 0.933 0.973 0.967 0.134 0.629
19 (7.1%)
3 (12.5%)
0.569
15 (5.6%)
2 (8.3%)
0.918
67 39 35 6 36 23
8 0 4 1 2 0
0.359 0.054 0.843 0.453 0.702 0.275
(24.8%) (14.4%) (13.0%) (2.2%) (13.3%) (8.5%)
(33.3%) (0%) (16.7%) (4.2%) (8.3%) (0%)
BSA ¼ body surface area; CAVC ¼ complete atrioventricular canal; DORV ¼ double outlets of the right ventricle; PAI¼pulmonary artery (Nakata) index; TA ¼ tricuspid atresia; TGA ¼ transposition of the great arteries.
Previous palliative procedures were performed in 11 patients; 10 patients underwent modified Blalock-Taussig shunt placement, and 1 patient underwent PAB. Pre-BDG cardiac catheterization was performed in all patients, and the pulmonary artery index (Nakata index) and mean pulmonary artery pressure were recorded. Forty-eight (16.3%) patients achieved TCPC. During 2012 to 2013, only 3 patients (3/103) who underwent BDG operations achieved TCPC. Therefore, the 151 patients who received BDG with uncontrolled AnPBF from 2008 to 2012 were further divided into two groups: group DF (delayed Fontan completion, n ¼ 109) and group FC (Fontan completion, n ¼ 42). The patients’ characteristics are summarized in Table 2.
Surgical Technique Before BDG, the major aortopulmonary collateral arteries were occluded (n ¼ 9) in a hybrid operating room. All the surgical procedures were performed through a median sternotomy. One hundred twenty-eight patients underwent BDG operations with the assistance of cardiopulmonary bypass. All previously placed systemic-topulmonary shunts were removed or ligated. For the BDG anastomosis, the SVC was anastomosed to the pulmonary artery branch in an end-to-side fashion. The following associated surgical procedures were performed: repair of total anomalous pulmonary venous drainage (TAPVC) (n ¼ 13), atrioventricular valvuloplasty (n ¼ 3), and pulmonary arterioplasty (n ¼ 21). The pulmonary artery pressures were monitored with transducers through the SVC after Glenn anastomosis. The main pulmonary artery was banded with a strip of polytetrafluoroethylene to achieve a mean pulmonary artery pressure or a CVP of less than 16 mm Hg. If the CVP was lower than 16 mm Hg, the azygos vein was ligated, and the main pulmonary artery was left intact. The azygos vein was left intact in group 2.
Follow-Up The mean follow-up time was 39 20 months (range, 5 to 78 months) for groups 1 and 2. The primary outcome was all-cause death, and the secondary outcome was TCPC completion. Early clinical outcomes included surgical mortality, pulmonary artery pressure, arterial oxygen saturation, and the amount of the pleural drainage before Table 2. Characteristics of Patients With and Without Fontan Completion
Age (years) Sex (M/F) Body weight (kg) BSA (m2) Oxygen saturation (%) PAI (mm2/m2)
Group DF n ¼ 109
Group FC n ¼ 42
p value
2.1 (0.4–19) 70/39 12.9 6.4 0.56 0.2 78.0 6.5 252.1 128.0
2.2 (0.4–25) 28/14 13.8 9.0 0.56 0.3 77.9 6.2 249.5 126.7
0.49 0.78 0.52 0.99 0.97 0.92
BSA ¼ body surface area; DF ¼ delayed Fontan; completion; PAI ¼ pulmonary artery (Nakata) index.
FC ¼ Fontan
Ann Thorac Surg 2016;-:-–-
ZHANG ET AL OUTCOMES OF BDG IN CHINA
Table 3. Multivariate Analysis of the Need for Pulmonary Artery Banding After the Bidirectional Glenn Procedure
SaO2 pre-Glenn SaO2 post-Glenn (pre-PAB) Associated procedures
Group 1 n ¼ 270
Group 2 n ¼ 24
p Value
OR
78.0 6.2 90.5 3.8
82.6 4.4 93.3 4.0
0.237 0.185
1.14 1.18
10 (3.7%)
6 (25%)
0.001
83.62
OR ¼ odds ratio; PAB ¼ pulmonary artery banding; arterial oxygen saturation.
SaO2¼
discharge. Echocardiographic studies were performed before discharge to assess ventricular function and atrioventricular valve regurgitation. The degree of valve regurgitation was graded from 0 to 4 using Sellers’ classification and is numerically expressed as follows: 0, none; 1, trivial; 2, mild; 3, moderate; and 4, severe [15]. Patients were encouraged to return for routine outpatient visits after hospital discharge. At each visit, body weight, arterial oxygen saturation, and echocardiographic data were recorded. Cardiac catheterizations were recommended for patients who were suitable for TCPC completion during follow-up. One hundred eighty-nine patients (64%) received subsequent cardiac catheterizations. The hemodynamic variables before BDG and Fontan completion are summarized in Table 5. Among these patients, 13 were excluded from Fontan completion because of inappropriate anatomy or hemodynamics (11 in group 1 and 2 in group 2).
Statistical Analysis The statistical calculations were performed using (SPSS Inc, Chicago, IL). The data are presented as the mean
3
standard deviation, and percentages are provided where appropriate. For continuous variables and comparisons of serial values, analysis of variance (ANOVA) was performed with the grouping variable followed by two-way ANOVA of the two separate groups. For the withingroup and between-group analyses, independent-sample t tests and c2 analyses were used, respectively. The Kaplan-Meier method and log-rank test were used for the actuarial survival analysis. Binary logistic regression analyses were performed to investigate the relationships of various independent variables with PAB and atrioventricular valve regurgitation aggravation. Statistical significance was defined by a p value of less than 0.05.
Results The amounts of pleural drainage before discharge were similar in groups 1 and 2 (114 68 mL versus 132 28 mL, p ¼ 0.48). Three early postoperative deaths resulted in a mortality rate of 0.8%. Two of these patients died of pulmonary hypertension and subsequent heart failure in group 1, and 1 patient died of severe pneumonia in group 2. Nine late deaths occurred during follow-up (3 patients in group DF and 0 in group FC). The actuarial survival rates did not significantly differ among the various groups (Fig 1). The mean pre-BDG pulmonary artery pressures were 15 3.7 mm Hg in group 1 and 19.1 5.9 mm Hg in group 2. The mean CVPs immediately after the Glenn anastomoses were 13.6 3.7 mm Hg in group 1 and 18.7 2.2 mm Hg in group 2. In group 2, the CVP decreased immediately after PAB (18.7 2.2 mm Hg versus 14.4 1.9 mm Hg, p < 0.001), but the arterial oxygen saturation was maintained (93.3 4.0% versus 89.6 3.4%, p ¼ 0.055). Logistic Fig 1. Estimated Kaplan-Meyer survival curves of the groups, including early mortality, after bidirectional Glenn anastomosis (p ¼ 0.51, Mantel-Cox).
4
ZHANG ET AL OUTCOMES OF BDG IN CHINA
regression analysis indicated that the surgical procedures (atrioventricular valvuloplasty or TAPVC repair) were associated with the need for PAB (Table 3). Mean postBDG CVPs before discharge were similar in group 1 (12.9 2.7 mm Hg) and group 2 (13.1 2.4 mm Hg; p ¼ 0.87). During follow-up, there were no differences in pulmonary artery pressures or Nakata indices after BDG between groups 1 and 2 (Fig 2). There were no withingroup or between-group differences in the systemic ventricular ejection fraction either before or after BDG (Fig 3A). In group 1, the systemic ventricular end-diastolic diameter increased from 25.5 8.5 mm to 29.8 10.8 mm (p ¼ 0.02, Fig 3B). Ventricular enlargement occurred in 66.7% of patients with right ventricular (RV) morphology, in 33.3% of patients with left ventricular morphology, and in 28.6% of patients with biventricular or uncertain morphology. The degrees of atrioventricular valve regurgitation were not significantly different before and after BDG in group 2. In group 1, the degree of atrioventricular valve regurgitation increased significantly from 1.6 1.1 (after BDG) to 2.1 0.8 (before TCPC, p ¼ 0.01, Fig 3C). Logistic regression analysis indicated that systemic RV morphology was a risk factor for the aggravation of valve regurgitation after BDG (odds ratio: 3.59, p ¼ 0.03; Table 4). There were no differences in the systemic ventricular end-diastolic pressure, right atrial pressure, or SVC saturation before BDG and Fontan completion between groups 1 and 2 (see Table 5). The average interval for TCPC completion was 2.2 1.1 years. There were no differences in the TCPC completion rates (15.9% versus 20.8%, p ¼ 0.42) or the interval between groups 1 and 2 (2.3 1.1 versus 1.8 1.0 years, p ¼ 0.38). The mean follow-up time was 48 15 months for groups DF and FC. At the end of the follow-up, all patients in group DF were still waiting for Fontan completion. The mean delay period was 48 months (range, 26 to 77 months). There was no difference in body weight gain between groups DF and FC (Fig 4A). In group FC, the majority of patients underwent TCPC at 24 months after BDG. Therefore, arterial oxygen saturation was significantly increased in group FC and slightly decreased in group DF at 24 months after BDG (Fig 4B). During the first 24 months after BDG, the end-diastolic diameter tended to increase in groups DF and FC. However, the tendency toward deterioration was worsened in group DF, and aggravation of valve regurgitation was observed during follow-up (see Figs 5B, 5C). There were no differences in ejection fractions during followup (see Fig 5A).
Comment BDG has been a standard intermediate palliative therapy for patients who are candidates for a subsequent Fontan operation [16]. In many centers in Western countries, the main pulmonary arteries are interrupted, and thus the SVC becomes the sole source of blood flow to the
Ann Thorac Surg 2016;-:-–-
Table 4. Multivariate Analysis for Atrioventricular Valve Regurgitation Aggravation in Group 1
RV morphology LV morphology Biventricular or uncertain morphology SVEDD pre-Glenn (mm) Associated procedures Degree moderate
AVVR Aggravation n ¼ 37
p Value
OR
10 (27%) 3 (8.1%) 4 (10.8%)
0.031 0.863 0.531
3.594 1.114 1.253
27 10.6 2 (5.4%) 4 (10.8%)
0.822 0.743 0.064
1.185 1.295 1.315
AVVR ¼ atrioventricular valve regurgitation; LV ¼ left ventricular; OR ¼ odds ratio; RV ¼ right ventricular; SVEDD ¼ systemic ventricular end-diastolic diameter.
Fig 2. Dynamic changes in mean pulmonary artery pressures and Nakata indices during follow-up. (A) Mean pulmonary artery pressure; *p < 0.01 for group 2 versus group 1. (B) The Nakata index (The post- bidirectional Glenn procedure (BDDG) mean pulmonary artery pressure was obtained from the central venous pressure after extubation. The pre-Fontan mean pulmonary artery pressure and Nakata index were recorded from the last cardiac catheterization).
Ann Thorac Surg 2016;-:-–-
Fig 3. Dynamic changes in ejection fraction, systemic ventricular end-diastolic diameter (SVEDD), and the degree of atrioventricular (AV) valve regurgitation during follow-up. (A) Ejection fraction. (B) SVEDD; *p ¼ 0.02 for group 1. (C) AV valve regurgitation; *p ¼ 0.01 for group 1 (the pre-Glenn, post-Glenn and pre-Fontan data were recorded from the last echocardiographic study before the Glenn procedure, after the Glenn procedure, and before the Fontan completion, respectively).
Table 5. Patients’ Hemodynamic Variables Before the Bidirectional Glenn Procedure and the Fontan Procedure pre-BDG Group 1 n ¼ 270
BDG ¼ bidirectional Glenn procedure.
78.0 4.8 9.8 66.3 15.0 7.8 54.5
6.2 1.2 2.2 18.4 3.7 2.2 4.5
Group 2 n ¼ 24
p value
0.002 0.324 0.049 0.429 0.038 0.263 0.425
82.6 5.3 12.5 65.7 19.1 8.3 58.3
4.4 1.3 3.2 31.9 5.9 1.8 5.2
Group 1 n ¼ 270 85.6 4.9 7.7 63.8 12.9 7.4 61.1
4.3 1.8 2.4 24.4 3.7 1.9 3.2
Group 2 n ¼ 24
p value
0.679 0.852 0.378 0.478 0.983 0.232 0.563
87.6 5.0 8.2 58.9 13.3 7.9 64.5
3.5 1.6 1.9 24.0 3.2 1.2 4.0
ZHANG ET AL OUTCOMES OF BDG IN CHINA
Systemic arterial saturation (%) Right atrial pressure (mm Hg) Transpulmonary gradient (mm Hg) Transpulmonary valve gradient (mm Hg) Pulmonary artery pressure (mm Hg) Ventricular end diastolic pressure (mm Hg) Superior vena cava saturation (%)
pre-Fontan
5
6
ZHANG ET AL OUTCOMES OF BDG IN CHINA
Ann Thorac Surg 2016;-:-–-
Fig 4. Changes in body weight and arterial oxygen saturation during follow-up. (A) Body weight. (B) Arterial oxygen saturation; *p < 0.05 for Group DF; **p < 0.05 for group DF versus group FC.
lungs [17]. However, the absence of pulsatile AnPBF could jeopardize pulmonary artery growth [7, 18]. In the present study, no tendency toward an increase in the Nakata index was observed after BDG. The mechanical stimuli caused by pulsatile AnPBF with the “hepatic factor” could contribute to pulmonary artery growth [19, 20]. In developed countries, the TCPC completion rate is 80% to 90%, and the interval between BDG and TCPC is
12 to 27 months [21–23]. In contrast, the Fontan completion rate in some Asian countries was as low as 15% in 2000 but has gradually improved to 40% to 50% in a few centers within the last decade. However, the interval time is still much longer (45 to 48 months) [12, 24, 25]. Therefore, maintaining the AnPBF during BDG has become a popular choice in China and India. Berdat and colleagues [3] reported that the maintenance or placement of
Ann Thorac Surg 2016;-:-–-
systemic-to-pulmonary shunts was a risk factor for early death after BDG. In our series, all previously existing systemic-to-pulmonary shunts, including Blalock-Taussig shunts, major aortopulmonary collateral arteries, and patent ductus arteriosus, were excluded before BDG. This policy maximally reduced the risk of elevations of pulmonary vascular resistance in patients with long waiting times for TCPC as indicated by the lack of observations of SVC syndrome and increasing pulmonary artery pressure. A theoretical analysis involving computational fluid dynamics revealed a linear relationship between additional PBF and CVP after BDG [26]. Yoshida and associates [18] advocated that PBF should be controlled and that an average CVP lower than 16 mm Hg is an appropriate threshold for guiding PAB. We applied the same threshold, and the CVP immediately decreased by 23% while the oxygen saturation remained stable. Associated surgical procedures (ie, atrioventricular valvuloplasty and TAPVC repair) strongly indicated the need for PAB after BDG. The patients with functional SVs and TAPVC had complicated anatomy and physiology. Postoperative pulmonary hypertension still exists because of poor development of the pulmonary vein in patients with TAPVC [27]. Therefore, BDG with controlled AnPBF could be the best choice and could become the definitive palliative therapy for patients. Additional PBF could result in volume overload of the functional SV and aggravation of valve regurgitation. Although no impaired systolic heart functions or elevated systemic ventricular end-diastolic pressures were observed, we did observe moderate ventricular dilatation and aggravation of valve regurgitation in the patients with uncontrolled AnPBF. Moreover, the logistic regression analysis revealed that RV morphology was strongly associated with this phenomenon. Atrioventricular valve insufficiency exerts a strong negative effect on Fontan completion [28]. The atrioventricular valves in patients with SV and RV morphology are less functional than those in patients with biventricular or left ventricular morphologies [29]. Therefore, for patients with RV morphologies, if the AnPBF is maintained without PAB because of a low CVP, then poor tolerance to the volume overload would contribute to the aggravation of valve regurgitation after BDG. In the present study, only 16.3% of patients achieved TCPC. Uncontrolled AnPBF produced higher arterial oxygen saturation, but declines in arterial oxygen saturation were observed in all patients during the first 24 months of follow-up. We did not observe the formation of any obvious pulmonary arteriovenous malformations depending on catheterization. The blood flow from the SVC to the pulmonary artery gradually decreases during a child’s development. Furthermore, the diminished AnPBF could also contribute to the decrease in arterial oxygen saturation. However, until now, we had no approach for accurately measuring the AnPBF during child development. The survival rates and body weight gains were similar between the patients with and without Fontan completion (see Figs 1, 4A). Although heart
ZHANG ET AL OUTCOMES OF BDG IN CHINA
Fig 5. Dynamic changes in ejection fraction (EF), systemic ventricular end-diastolic diameter (SVEDD), and the degree of atrioventricular valve regurgitation during follow-up. (A) Ejection fraction. (B) SVEDD. (C) Atrioventricular valve regurgitation; *p < 0.05 for group DF, **p > 0.05 for group FC.
7
8
ZHANG ET AL OUTCOMES OF BDG IN CHINA
function was preserved (Fig 5A), uncontrolled AnPBF will eventually lead to the enlargement of univentricular hearts and the aggravation of valve regurgitation (Figs 5B, 5C); therefore, this situation could make these patients unsuitable for Fontan completion. Congenital heart diseases have become the most frequent birth defects in China [30]. However, the low Fontan achievement rate and the long interval represent real-world outcomes in China. With improvements in the public health insurance system, more patients will have the opportunity to complete TCPC within an appropriate waiting time. However, in the current conditions, we have to apply an aggressive strategy of maintaining the AnPBF and excluding the use of any additional systemic-to-pulmonary shunts. Because of the limited case numbers, we did not evaluate the effect of controlled AnPBF on patients with delayed Fontan completion. In the present study, PAB was performed based on the presence of high CVP (or high pulmonary artery pressure) after BDG. During child development, PAB could restrict excessive AnPBF. Hence, even if the CVP after BDG is acceptable in the operating room, PAB can be performed to control AnPBF, particularly in patients with univentricular hearts and RV morphologies. The use of a pulmonary artery pressure threshold after BDG to guide PAB for the restriction of AnPBF requires further investigation. In conclusion, the low Fontan achievement rate is a critical issue in China. The aggressive policy of maintaining AnPBF after BDG provided acceptable arterial oxygen saturations and an acceptable survival rate for patients with delayed Fontan completion. However, uncontrolled AnPBF was associated with ventricular enlargement and aggravation of valve regurgitation. Strategies to improve the Fontan completion rate in China should be explored and could improve the outcomes. This study was supported by the Program for New Century Excellent Talents in University and the Foundation for Young Investigators at Peking Union Medical College (XHQN09 and XR01-YP). The authors wish to thank Dr. Ming-Sing Si from the Department of Cardiac Surgery, University of Michigan for his language editing assistance.
References 1. Tanoue Y, Sese A, Imoto Y, Joh K. Ventricular mechanics in the bidirectional Glenn procedure and total cavopulmonary connection. Ann Thorac Surg 2003;76:562–6. 2. Alphonso N, Baghai M, Sundar P, Tulloh R, Austin C, Anderson D. Intermediate-term outcome following the Fontan operation: a survival functional and risk-factor analysis. Eur J Cardiothorac Surg 2005;28:529–35. 3. Berdat PA, Belli E, Lacour-Gayet F, Planche C, Serraf A. Additional pulmonary blood flow has no adverse effect on outcome after bidirectional cavopulmonary shunt. Ann Thorac Surg 2005;79:29–36. 4. Fiore AC, Tobin C, Jureidini S, Rahimi M, Kim ES, Schowengerdt K. A comparison of the modified BlalockTaussig shunt with the right ventricle-to-pulmonary artery conduit. Ann Thorac Surg 2011;91:1479–85.
Ann Thorac Surg 2016;-:-–-
5. Gray RG, Altmann K, Mosca RS, et al. Persistent antegrade pulmonary blood flow post-Glenn does not alter early postFontan outcomes in single-ventricle patients. Ann Thorac Surg 2007;84:888–93. 6. Gerelli S, Boulitrop C, Steenberghe MV, et al. Bidirectional cavopulmonary shunt with additional pulmonary blood flow: a failed or successful strategy? Eur J Cardiothorac Surg 2012;42:513–9. 7. Caspi J, Pettitt TW, Ferguson TB, Stopa AR, Sandhu SK. Effects of controlled antegrade pulmonary blood flow on cardiac function after bidirectional cavopulmonary anastomosis. Ann Thorac Surg 2003;76:1917–22. 8. Calvaruso DF, Rubino A, Ocello S, et al. Bidirectional Glenn and antegrade pulmonary blood flow: temporary or definitive palliation? Ann Thorac Surg 2008;85:1389–95. 9. van Slooten Y, Elzenga NJ, Waterbolk TW, van Melle JP, Berger RMF, Ebels T. The effect of additional pulmonary blood flow on timing of the total cavopulmonary connection. Ann Thorac Surg 2012;93:2028–34. 10. McElhinney DB, Marianeschi SM, Reddy VM. Additional pulmonary blood flow with the bidirectional Glenn anastomosis: does it make a difference? Ann Thorac Surg 1998;66: 668–72. 11. Ferns SJ, Zein CE, Multani K, et al. Is additional pulsatile pulmonary blood flow beneficial to patients with bidirectional Glenn? J Thorac Cardiovasc Surg 2013;145:451–4. 12. Xie B, Zhang JF, Shetty DP. Bidirectional Glenn shunt: 170 cases. Asian Cardiovasc Thorac Ann 2001;9:196–9. 13. Zhang WM, Zhang J, Wu ZK, Bai J. Abudushalamu. Bidirectional Glenn procedure for the treatment of complex cyanotic congenital heart diseases. Chin J Clin Thorac Cardiovasc Surg 2014;21:482–5. 14. Pankaj G, Kona SM, Kotturathu MC. Bidirectional cavopulmonary shunt with accessory pulmonary blood flow. Asian Cardiovasc Thorac Ann 2001;9:279–81. 15. Yuji N, Takeshi H. Long-term results of modified Fontan operation for single ventricle patients associated with atrioventricular valve regurgitation. Ann Thorac Surg 2013;96: 211–8. 16. Schrurer MA, Hill EG, Vasuki N, et al. Survival after bidirectional cavopulmonary anastomosis: analysis of preoperative risk factors. J Thorac Cardiovasc Surg 2007;134:82–9. 17. Schreiber C, Cleuziou J, Cornelsen JK, Horer J, Eicken A, Lange R. Bidirectional cavopulmonary connection without additional pulmonary blood flow as an ideal staging for functional univentricular hearts. Eur J Cardiothorac Surg 2008;34:550–5. 18. Yoshida M, Yamaguchi M, Yoshimura N, Murakami H, Matsuhisa H, Okita Y. Appropriate additional pulmonary blood flow at the bidirectional Glenn procedure is useful for completion of total cavopulmonary connection. Ann Thorac Surg 2005;80:976–81. 19. Vettukattil JJ, Slavik Z, Lamb RK, et al. Intrapulmonary arteriovenous shunting may be universal phenomenon in patients with the superior cavopulmonary anastomosis: a radionuclide study. Heart 2000;83:425–8. 20. Kurotobi S, Sano T, Kogaki S, et al. Bidirectional cavopulmonary shunt with right ventricular outflow patency: the impact of pulsatility on pulmonary endothelial function. J Thorac Cardiovasc Surg 2001;121:1161–8. 21. Tanoue Y, Kado H, Boku N, et al. Three hundred and thirtythree experiences with the bidirectional Glenn procedure in a single institute. Interact Cardiovasc Thorac Surg 2007;6: 97–101. 22. Rogers LS, Glatz AC, Ravishankar C, et al. 18 years of the Fontan operation at single institution. J Am Coll Cardiol 2012;60:1018–25. 23. Tan AM, Iyengar AJ, Donath S, et al. Fontan completion rate and outcomes after bidirectional cavopulmonary shunt. Eur J Cardiothorac Surg 2010;38:59–65. 24. Ganigara M, Parbhu A, Varghese R, Pavithran S, Valliatu J, Nair R. Extracardiac Fontan operation after late
Ann Thorac Surg 2016;-:-–-
bidirectional Glenn shunt. Asian Cardiovasc Thorac Ann 2010;18:253–9. 25. Suchaya S, Yupada P. Factors influencing survival in patients after bidirectional Glenn shunt. Asian Cardiovasc Thorac Ann 2008;16:381–6. 26. Sun Q, Wan D, Liu J, et al. Influence of antegrade pulmonary blood flow on the hemodynamic performance of bidirectional cavopulmonary anastomosis: a numerical study. Med Eng Phys 2009;31:227–33. 27. Lodge AJ, Rjchik J, Nicolson SC, Ittenbach RF, Spary TL, Gaynor JW. Improving outcomes in functional single
ZHANG ET AL OUTCOMES OF BDG IN CHINA
9
ventricle and total anomalous pulmonary venous connection. Ann Thorac Surg 2004;78:1688–95. 28. Lee TM, Aiyagari R, Hirsch JC, Ohye RG, Bove EL, Devaney EJ. Risk factor analysis for second-stage palliation of single ventricle anatomy. Ann Thorac Surg 2012;93:614–9. 29. Khairy P, Poirier N, Mercier LA. Univentricular heart. Circulation 2007;115:800–12. 30. Hu Z, Yuan X, Rao K, Zheng Z, Hu S. National trend in congenital heart disease mortality in China during 2003 to 2010: a population-based study. J Thorac Cardiovasc Surg 2014;148:596–602.
Background. Given the low rate of Fontan completion, an aggressive policy for maintaining antegrade pulmonary blood flow (AnPBF) during the bidirectional Glenn procedure (BDG) was developed for the functional single ventricle. Methods. From 2008 to 2013, 294 patients who underwent the BDG were divided into two groups: group 1 (uncontrolled AnPBF, n [ 270) and group 2 (controlled AnPBF, n [ 24). Pulmonary artery banding was performed because of the high central venous pressure in group 2. In group 1, the patients who underwent BDG from 2008 to 2012 were further divided into group DF (delayed Fontan completion, n [ 109) and group FC (Fontan completion, n [ 42). Results. The Fontan completion rate was 16.3%, and the average interval time was 2.2 ± 1.1 years. The delay of Fontan completion did not reduce body weight gain or the survival rate. Furthermore, oxygen saturation was slightly reduced in group DF. Although no impairments
of heart function were observed, the uncontrolled AnPBF in group DF resulted in an increase in ventricular enddiastolic diameter and aggravation of atrioventricular valve regurgitation over 24 months after BDG. Logistic regression analysis revealed that systemic right ventricular morphology was a risk factor for the aggravation of valve regurgitation. Conclusions. The low Fontan achievement rate is a critical issue in China. Although the patients with delayed Fontan completion exhibited an acceptable survival rate and acceptable body weight gain, uncontrolled AnPBF was associated with ventricular enlargement and aggravation of valve regurgitation. Strategies for improving the Fontan completion rate in China should be explored and could benefit outcomes.
T
the superior vena cava (SVC) pressure and result in the persistent pleural effusions and functional SV volume overload [10, 11]. The proportion of patients who eventually complete TCPC in some Asian countries is lower than that in developed Western countries. A group from India reported on 170 patients who underwent BDG operations, and only 15% of these patients received subsequent TCPC after an average interval of 48 months [12]. Most recently, a heart center located in a relatively underdeveloped area in Northern China reported the results of 68 BDG cases, and the rate of TCPC achievement was as low as 5.8% (4/68) after an interval of 3 to 5 years [13]. Low economic status and undeveloped insurance systems have contributed to the dismal Fontan completion rates. Hence, AnPBF has been maintained in many heart centers in China and India to increase and maintain arterial oxygen saturation after BDG [14]. This strategy may provide more time for the patient’s family to realize the importance of subsequent treatment and to raise the funds necessary for TCPC. Therefore, we apply an aggressive policy of maintaining AnPBF without any systemic pulmonary shunts. Here, we performed a
he bidirectional Glenn procedure (BDG) has become an effective palliative procedure and a standard intermediate step toward total cavopulmonary connection (TCPC) for patients with a functional single ventricle (SV) [1, 2]. The main sources of additional pulmonary blood flow (PBF) after BDG include systemic-to-pulmonary shunts and antegrade PBF (AnPBF). Systemic-topulmonary shunts may increase perioperative mortality, negatively affect left pulmonary artery growth, and potentially increase pulmonary vascular resistance [3, 4]. Therefore, the placement of systemic-to-pulmonary shunts is rarely advocated for patients with hypoxia after BDG. Controversy still exists regarding maintenance of AnPBF after BDG [5, 6]. Pulsatile AnPBF may improve arterial oxygen saturation, stimulate pulmonary artery growth, and prevent the formation of arteriovenous fistulas [7–9]. However, pulsatile AnPBF may also increase Accepted for publication Oct 26, 2015. Address correspondence to Dr Hao Zhang, Center for Pediatric Cardiac Surgery, Fuwai Hospital, Chinese Academy of Medical Sciences, Beijing, 100037, China; email: [email protected].
Ó 2016 by The Society of Thoracic Surgeons Published by Elsevier
(Ann Thorac Surg 2016;-:-–-) Ó 2016 by The Society of Thoracic Surgeons
0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2015.10.071
2
ZHANG ET AL OUTCOMES OF BDG IN CHINA
Ann Thorac Surg 2016;-:-–-
retrospective analysis of the patients with functional SV who were treated at our center. In this study, we analyzed the midterm clinical outcomes of patients with and without Fontan completion. We sought to evaluate the outcomes of this aggressive policy of maintaining AnPBF, to address the potential beneficial effects of controlled AnPBF for patients waiting for TCPC, and to emphasize the influence of uncontrolled AnPBF on patients with delayed TCPC.
Material and Methods Study Population The ethics committee of Fuwai Hospital in Beijing approved this study. We retrospectively reviewed the clinical records of 420 patients without extracardiac anomalies or chromosomal abnormalities who underwent BDG from January 2008 to December 2013. Fortyseven patients who underwent one and one-half ventricle repair, 5 patients who received modified Blalock-Taussig shunt placement after BDG, and 74 patients with pulmonary atresia were excluded. The remaining patients with SV physiology were divided into two groups: group 1 included the patients with uncontrolled AnPBF (n ¼ 270), and group 2 included the patients in whom AnPBF was controlled by pulmonary artery banding (PAB) as a result of high SVC pressure (or central venous pressure, CVP) (n ¼ 24). The patients’ characteristics are summarized in Table 1. There were no significant differences in age, sex, body weight, body surface area, or the Nakata index between groups 1 and 2. Table 1. Patient Characteristics and Anatomic Lesions at the Time of the Bidirectional Glenn Procedure
Age (years) Sex (/F) Body weight (kg) BSA (m2) PAI (mm2/m2) Single ventricle Left ventricle morphology Right ventricle morphology Biventricular or uncertain DORV TA TGA Ebstein anomaly CAVC Others
Group 1 n ¼ 270
Group 2 n ¼ 24
p value
2 (0.3–25) 173/97 13.6 9.3 0.6 0.3 252.3 125.4 64 (23.8%) 30 (11.1%)
2.1 (0.5–11) 15/9 14.7 7.9 0.6 0.2 225.5 86.5 9 (37.5%) 4 (16.7%)
0.563 0.822 0.933 0.973 0.967 0.134 0.629
19 (7.1%)
3 (12.5%)
0.569
15 (5.6%)
2 (8.3%)
0.918
67 39 35 6 36 23
8 0 4 1 2 0
0.359 0.054 0.843 0.453 0.702 0.275
(24.8%) (14.4%) (13.0%) (2.2%) (13.3%) (8.5%)
(33.3%) (0%) (16.7%) (4.2%) (8.3%) (0%)
BSA ¼ body surface area; CAVC ¼ complete atrioventricular canal; DORV ¼ double outlets of the right ventricle; PAI¼pulmonary artery (Nakata) index; TA ¼ tricuspid atresia; TGA ¼ transposition of the great arteries.
Previous palliative procedures were performed in 11 patients; 10 patients underwent modified Blalock-Taussig shunt placement, and 1 patient underwent PAB. Pre-BDG cardiac catheterization was performed in all patients, and the pulmonary artery index (Nakata index) and mean pulmonary artery pressure were recorded. Forty-eight (16.3%) patients achieved TCPC. During 2012 to 2013, only 3 patients (3/103) who underwent BDG operations achieved TCPC. Therefore, the 151 patients who received BDG with uncontrolled AnPBF from 2008 to 2012 were further divided into two groups: group DF (delayed Fontan completion, n ¼ 109) and group FC (Fontan completion, n ¼ 42). The patients’ characteristics are summarized in Table 2.
Surgical Technique Before BDG, the major aortopulmonary collateral arteries were occluded (n ¼ 9) in a hybrid operating room. All the surgical procedures were performed through a median sternotomy. One hundred twenty-eight patients underwent BDG operations with the assistance of cardiopulmonary bypass. All previously placed systemic-topulmonary shunts were removed or ligated. For the BDG anastomosis, the SVC was anastomosed to the pulmonary artery branch in an end-to-side fashion. The following associated surgical procedures were performed: repair of total anomalous pulmonary venous drainage (TAPVC) (n ¼ 13), atrioventricular valvuloplasty (n ¼ 3), and pulmonary arterioplasty (n ¼ 21). The pulmonary artery pressures were monitored with transducers through the SVC after Glenn anastomosis. The main pulmonary artery was banded with a strip of polytetrafluoroethylene to achieve a mean pulmonary artery pressure or a CVP of less than 16 mm Hg. If the CVP was lower than 16 mm Hg, the azygos vein was ligated, and the main pulmonary artery was left intact. The azygos vein was left intact in group 2.
Follow-Up The mean follow-up time was 39 20 months (range, 5 to 78 months) for groups 1 and 2. The primary outcome was all-cause death, and the secondary outcome was TCPC completion. Early clinical outcomes included surgical mortality, pulmonary artery pressure, arterial oxygen saturation, and the amount of the pleural drainage before Table 2. Characteristics of Patients With and Without Fontan Completion
Age (years) Sex (M/F) Body weight (kg) BSA (m2) Oxygen saturation (%) PAI (mm2/m2)
Group DF n ¼ 109
Group FC n ¼ 42
p value
2.1 (0.4–19) 70/39 12.9 6.4 0.56 0.2 78.0 6.5 252.1 128.0
2.2 (0.4–25) 28/14 13.8 9.0 0.56 0.3 77.9 6.2 249.5 126.7
0.49 0.78 0.52 0.99 0.97 0.92
BSA ¼ body surface area; DF ¼ delayed Fontan; completion; PAI ¼ pulmonary artery (Nakata) index.
FC ¼ Fontan
Ann Thorac Surg 2016;-:-–-
ZHANG ET AL OUTCOMES OF BDG IN CHINA
Table 3. Multivariate Analysis of the Need for Pulmonary Artery Banding After the Bidirectional Glenn Procedure
SaO2 pre-Glenn SaO2 post-Glenn (pre-PAB) Associated procedures
Group 1 n ¼ 270
Group 2 n ¼ 24
p Value
OR
78.0 6.2 90.5 3.8
82.6 4.4 93.3 4.0
0.237 0.185
1.14 1.18
10 (3.7%)
6 (25%)
0.001
83.62
OR ¼ odds ratio; PAB ¼ pulmonary artery banding; arterial oxygen saturation.
SaO2¼
discharge. Echocardiographic studies were performed before discharge to assess ventricular function and atrioventricular valve regurgitation. The degree of valve regurgitation was graded from 0 to 4 using Sellers’ classification and is numerically expressed as follows: 0, none; 1, trivial; 2, mild; 3, moderate; and 4, severe [15]. Patients were encouraged to return for routine outpatient visits after hospital discharge. At each visit, body weight, arterial oxygen saturation, and echocardiographic data were recorded. Cardiac catheterizations were recommended for patients who were suitable for TCPC completion during follow-up. One hundred eighty-nine patients (64%) received subsequent cardiac catheterizations. The hemodynamic variables before BDG and Fontan completion are summarized in Table 5. Among these patients, 13 were excluded from Fontan completion because of inappropriate anatomy or hemodynamics (11 in group 1 and 2 in group 2).
Statistical Analysis The statistical calculations were performed using (SPSS Inc, Chicago, IL). The data are presented as the mean
3
standard deviation, and percentages are provided where appropriate. For continuous variables and comparisons of serial values, analysis of variance (ANOVA) was performed with the grouping variable followed by two-way ANOVA of the two separate groups. For the withingroup and between-group analyses, independent-sample t tests and c2 analyses were used, respectively. The Kaplan-Meier method and log-rank test were used for the actuarial survival analysis. Binary logistic regression analyses were performed to investigate the relationships of various independent variables with PAB and atrioventricular valve regurgitation aggravation. Statistical significance was defined by a p value of less than 0.05.
Results The amounts of pleural drainage before discharge were similar in groups 1 and 2 (114 68 mL versus 132 28 mL, p ¼ 0.48). Three early postoperative deaths resulted in a mortality rate of 0.8%. Two of these patients died of pulmonary hypertension and subsequent heart failure in group 1, and 1 patient died of severe pneumonia in group 2. Nine late deaths occurred during follow-up (3 patients in group DF and 0 in group FC). The actuarial survival rates did not significantly differ among the various groups (Fig 1). The mean pre-BDG pulmonary artery pressures were 15 3.7 mm Hg in group 1 and 19.1 5.9 mm Hg in group 2. The mean CVPs immediately after the Glenn anastomoses were 13.6 3.7 mm Hg in group 1 and 18.7 2.2 mm Hg in group 2. In group 2, the CVP decreased immediately after PAB (18.7 2.2 mm Hg versus 14.4 1.9 mm Hg, p < 0.001), but the arterial oxygen saturation was maintained (93.3 4.0% versus 89.6 3.4%, p ¼ 0.055). Logistic Fig 1. Estimated Kaplan-Meyer survival curves of the groups, including early mortality, after bidirectional Glenn anastomosis (p ¼ 0.51, Mantel-Cox).
4
ZHANG ET AL OUTCOMES OF BDG IN CHINA
regression analysis indicated that the surgical procedures (atrioventricular valvuloplasty or TAPVC repair) were associated with the need for PAB (Table 3). Mean postBDG CVPs before discharge were similar in group 1 (12.9 2.7 mm Hg) and group 2 (13.1 2.4 mm Hg; p ¼ 0.87). During follow-up, there were no differences in pulmonary artery pressures or Nakata indices after BDG between groups 1 and 2 (Fig 2). There were no withingroup or between-group differences in the systemic ventricular ejection fraction either before or after BDG (Fig 3A). In group 1, the systemic ventricular end-diastolic diameter increased from 25.5 8.5 mm to 29.8 10.8 mm (p ¼ 0.02, Fig 3B). Ventricular enlargement occurred in 66.7% of patients with right ventricular (RV) morphology, in 33.3% of patients with left ventricular morphology, and in 28.6% of patients with biventricular or uncertain morphology. The degrees of atrioventricular valve regurgitation were not significantly different before and after BDG in group 2. In group 1, the degree of atrioventricular valve regurgitation increased significantly from 1.6 1.1 (after BDG) to 2.1 0.8 (before TCPC, p ¼ 0.01, Fig 3C). Logistic regression analysis indicated that systemic RV morphology was a risk factor for the aggravation of valve regurgitation after BDG (odds ratio: 3.59, p ¼ 0.03; Table 4). There were no differences in the systemic ventricular end-diastolic pressure, right atrial pressure, or SVC saturation before BDG and Fontan completion between groups 1 and 2 (see Table 5). The average interval for TCPC completion was 2.2 1.1 years. There were no differences in the TCPC completion rates (15.9% versus 20.8%, p ¼ 0.42) or the interval between groups 1 and 2 (2.3 1.1 versus 1.8 1.0 years, p ¼ 0.38). The mean follow-up time was 48 15 months for groups DF and FC. At the end of the follow-up, all patients in group DF were still waiting for Fontan completion. The mean delay period was 48 months (range, 26 to 77 months). There was no difference in body weight gain between groups DF and FC (Fig 4A). In group FC, the majority of patients underwent TCPC at 24 months after BDG. Therefore, arterial oxygen saturation was significantly increased in group FC and slightly decreased in group DF at 24 months after BDG (Fig 4B). During the first 24 months after BDG, the end-diastolic diameter tended to increase in groups DF and FC. However, the tendency toward deterioration was worsened in group DF, and aggravation of valve regurgitation was observed during follow-up (see Figs 5B, 5C). There were no differences in ejection fractions during followup (see Fig 5A).
Comment BDG has been a standard intermediate palliative therapy for patients who are candidates for a subsequent Fontan operation [16]. In many centers in Western countries, the main pulmonary arteries are interrupted, and thus the SVC becomes the sole source of blood flow to the
Ann Thorac Surg 2016;-:-–-
Table 4. Multivariate Analysis for Atrioventricular Valve Regurgitation Aggravation in Group 1
RV morphology LV morphology Biventricular or uncertain morphology SVEDD pre-Glenn (mm) Associated procedures Degree moderate
AVVR Aggravation n ¼ 37
p Value
OR
10 (27%) 3 (8.1%) 4 (10.8%)
0.031 0.863 0.531
3.594 1.114 1.253
27 10.6 2 (5.4%) 4 (10.8%)
0.822 0.743 0.064
1.185 1.295 1.315
AVVR ¼ atrioventricular valve regurgitation; LV ¼ left ventricular; OR ¼ odds ratio; RV ¼ right ventricular; SVEDD ¼ systemic ventricular end-diastolic diameter.
Fig 2. Dynamic changes in mean pulmonary artery pressures and Nakata indices during follow-up. (A) Mean pulmonary artery pressure; *p < 0.01 for group 2 versus group 1. (B) The Nakata index (The post- bidirectional Glenn procedure (BDDG) mean pulmonary artery pressure was obtained from the central venous pressure after extubation. The pre-Fontan mean pulmonary artery pressure and Nakata index were recorded from the last cardiac catheterization).
Ann Thorac Surg 2016;-:-–-
Fig 3. Dynamic changes in ejection fraction, systemic ventricular end-diastolic diameter (SVEDD), and the degree of atrioventricular (AV) valve regurgitation during follow-up. (A) Ejection fraction. (B) SVEDD; *p ¼ 0.02 for group 1. (C) AV valve regurgitation; *p ¼ 0.01 for group 1 (the pre-Glenn, post-Glenn and pre-Fontan data were recorded from the last echocardiographic study before the Glenn procedure, after the Glenn procedure, and before the Fontan completion, respectively).
Table 5. Patients’ Hemodynamic Variables Before the Bidirectional Glenn Procedure and the Fontan Procedure pre-BDG Group 1 n ¼ 270
BDG ¼ bidirectional Glenn procedure.
78.0 4.8 9.8 66.3 15.0 7.8 54.5
6.2 1.2 2.2 18.4 3.7 2.2 4.5
Group 2 n ¼ 24
p value
0.002 0.324 0.049 0.429 0.038 0.263 0.425
82.6 5.3 12.5 65.7 19.1 8.3 58.3
4.4 1.3 3.2 31.9 5.9 1.8 5.2
Group 1 n ¼ 270 85.6 4.9 7.7 63.8 12.9 7.4 61.1
4.3 1.8 2.4 24.4 3.7 1.9 3.2
Group 2 n ¼ 24
p value
0.679 0.852 0.378 0.478 0.983 0.232 0.563
87.6 5.0 8.2 58.9 13.3 7.9 64.5
3.5 1.6 1.9 24.0 3.2 1.2 4.0
ZHANG ET AL OUTCOMES OF BDG IN CHINA
Systemic arterial saturation (%) Right atrial pressure (mm Hg) Transpulmonary gradient (mm Hg) Transpulmonary valve gradient (mm Hg) Pulmonary artery pressure (mm Hg) Ventricular end diastolic pressure (mm Hg) Superior vena cava saturation (%)
pre-Fontan
5
6
ZHANG ET AL OUTCOMES OF BDG IN CHINA
Ann Thorac Surg 2016;-:-–-
Fig 4. Changes in body weight and arterial oxygen saturation during follow-up. (A) Body weight. (B) Arterial oxygen saturation; *p < 0.05 for Group DF; **p < 0.05 for group DF versus group FC.
lungs [17]. However, the absence of pulsatile AnPBF could jeopardize pulmonary artery growth [7, 18]. In the present study, no tendency toward an increase in the Nakata index was observed after BDG. The mechanical stimuli caused by pulsatile AnPBF with the “hepatic factor” could contribute to pulmonary artery growth [19, 20]. In developed countries, the TCPC completion rate is 80% to 90%, and the interval between BDG and TCPC is
12 to 27 months [21–23]. In contrast, the Fontan completion rate in some Asian countries was as low as 15% in 2000 but has gradually improved to 40% to 50% in a few centers within the last decade. However, the interval time is still much longer (45 to 48 months) [12, 24, 25]. Therefore, maintaining the AnPBF during BDG has become a popular choice in China and India. Berdat and colleagues [3] reported that the maintenance or placement of
Ann Thorac Surg 2016;-:-–-
systemic-to-pulmonary shunts was a risk factor for early death after BDG. In our series, all previously existing systemic-to-pulmonary shunts, including Blalock-Taussig shunts, major aortopulmonary collateral arteries, and patent ductus arteriosus, were excluded before BDG. This policy maximally reduced the risk of elevations of pulmonary vascular resistance in patients with long waiting times for TCPC as indicated by the lack of observations of SVC syndrome and increasing pulmonary artery pressure. A theoretical analysis involving computational fluid dynamics revealed a linear relationship between additional PBF and CVP after BDG [26]. Yoshida and associates [18] advocated that PBF should be controlled and that an average CVP lower than 16 mm Hg is an appropriate threshold for guiding PAB. We applied the same threshold, and the CVP immediately decreased by 23% while the oxygen saturation remained stable. Associated surgical procedures (ie, atrioventricular valvuloplasty and TAPVC repair) strongly indicated the need for PAB after BDG. The patients with functional SVs and TAPVC had complicated anatomy and physiology. Postoperative pulmonary hypertension still exists because of poor development of the pulmonary vein in patients with TAPVC [27]. Therefore, BDG with controlled AnPBF could be the best choice and could become the definitive palliative therapy for patients. Additional PBF could result in volume overload of the functional SV and aggravation of valve regurgitation. Although no impaired systolic heart functions or elevated systemic ventricular end-diastolic pressures were observed, we did observe moderate ventricular dilatation and aggravation of valve regurgitation in the patients with uncontrolled AnPBF. Moreover, the logistic regression analysis revealed that RV morphology was strongly associated with this phenomenon. Atrioventricular valve insufficiency exerts a strong negative effect on Fontan completion [28]. The atrioventricular valves in patients with SV and RV morphology are less functional than those in patients with biventricular or left ventricular morphologies [29]. Therefore, for patients with RV morphologies, if the AnPBF is maintained without PAB because of a low CVP, then poor tolerance to the volume overload would contribute to the aggravation of valve regurgitation after BDG. In the present study, only 16.3% of patients achieved TCPC. Uncontrolled AnPBF produced higher arterial oxygen saturation, but declines in arterial oxygen saturation were observed in all patients during the first 24 months of follow-up. We did not observe the formation of any obvious pulmonary arteriovenous malformations depending on catheterization. The blood flow from the SVC to the pulmonary artery gradually decreases during a child’s development. Furthermore, the diminished AnPBF could also contribute to the decrease in arterial oxygen saturation. However, until now, we had no approach for accurately measuring the AnPBF during child development. The survival rates and body weight gains were similar between the patients with and without Fontan completion (see Figs 1, 4A). Although heart
ZHANG ET AL OUTCOMES OF BDG IN CHINA
Fig 5. Dynamic changes in ejection fraction (EF), systemic ventricular end-diastolic diameter (SVEDD), and the degree of atrioventricular valve regurgitation during follow-up. (A) Ejection fraction. (B) SVEDD. (C) Atrioventricular valve regurgitation; *p < 0.05 for group DF, **p > 0.05 for group FC.
7
8
ZHANG ET AL OUTCOMES OF BDG IN CHINA
function was preserved (Fig 5A), uncontrolled AnPBF will eventually lead to the enlargement of univentricular hearts and the aggravation of valve regurgitation (Figs 5B, 5C); therefore, this situation could make these patients unsuitable for Fontan completion. Congenital heart diseases have become the most frequent birth defects in China [30]. However, the low Fontan achievement rate and the long interval represent real-world outcomes in China. With improvements in the public health insurance system, more patients will have the opportunity to complete TCPC within an appropriate waiting time. However, in the current conditions, we have to apply an aggressive strategy of maintaining the AnPBF and excluding the use of any additional systemic-to-pulmonary shunts. Because of the limited case numbers, we did not evaluate the effect of controlled AnPBF on patients with delayed Fontan completion. In the present study, PAB was performed based on the presence of high CVP (or high pulmonary artery pressure) after BDG. During child development, PAB could restrict excessive AnPBF. Hence, even if the CVP after BDG is acceptable in the operating room, PAB can be performed to control AnPBF, particularly in patients with univentricular hearts and RV morphologies. The use of a pulmonary artery pressure threshold after BDG to guide PAB for the restriction of AnPBF requires further investigation. In conclusion, the low Fontan achievement rate is a critical issue in China. The aggressive policy of maintaining AnPBF after BDG provided acceptable arterial oxygen saturations and an acceptable survival rate for patients with delayed Fontan completion. However, uncontrolled AnPBF was associated with ventricular enlargement and aggravation of valve regurgitation. Strategies to improve the Fontan completion rate in China should be explored and could improve the outcomes. This study was supported by the Program for New Century Excellent Talents in University and the Foundation for Young Investigators at Peking Union Medical College (XHQN09 and XR01-YP). The authors wish to thank Dr. Ming-Sing Si from the Department of Cardiac Surgery, University of Michigan for his language editing assistance.
References 1. Tanoue Y, Sese A, Imoto Y, Joh K. Ventricular mechanics in the bidirectional Glenn procedure and total cavopulmonary connection. Ann Thorac Surg 2003;76:562–6. 2. Alphonso N, Baghai M, Sundar P, Tulloh R, Austin C, Anderson D. Intermediate-term outcome following the Fontan operation: a survival functional and risk-factor analysis. Eur J Cardiothorac Surg 2005;28:529–35. 3. Berdat PA, Belli E, Lacour-Gayet F, Planche C, Serraf A. Additional pulmonary blood flow has no adverse effect on outcome after bidirectional cavopulmonary shunt. Ann Thorac Surg 2005;79:29–36. 4. Fiore AC, Tobin C, Jureidini S, Rahimi M, Kim ES, Schowengerdt K. A comparison of the modified BlalockTaussig shunt with the right ventricle-to-pulmonary artery conduit. Ann Thorac Surg 2011;91:1479–85.
Ann Thorac Surg 2016;-:-–-
5. Gray RG, Altmann K, Mosca RS, et al. Persistent antegrade pulmonary blood flow post-Glenn does not alter early postFontan outcomes in single-ventricle patients. Ann Thorac Surg 2007;84:888–93. 6. Gerelli S, Boulitrop C, Steenberghe MV, et al. Bidirectional cavopulmonary shunt with additional pulmonary blood flow: a failed or successful strategy? Eur J Cardiothorac Surg 2012;42:513–9. 7. Caspi J, Pettitt TW, Ferguson TB, Stopa AR, Sandhu SK. Effects of controlled antegrade pulmonary blood flow on cardiac function after bidirectional cavopulmonary anastomosis. Ann Thorac Surg 2003;76:1917–22. 8. Calvaruso DF, Rubino A, Ocello S, et al. Bidirectional Glenn and antegrade pulmonary blood flow: temporary or definitive palliation? Ann Thorac Surg 2008;85:1389–95. 9. van Slooten Y, Elzenga NJ, Waterbolk TW, van Melle JP, Berger RMF, Ebels T. The effect of additional pulmonary blood flow on timing of the total cavopulmonary connection. Ann Thorac Surg 2012;93:2028–34. 10. McElhinney DB, Marianeschi SM, Reddy VM. Additional pulmonary blood flow with the bidirectional Glenn anastomosis: does it make a difference? Ann Thorac Surg 1998;66: 668–72. 11. Ferns SJ, Zein CE, Multani K, et al. Is additional pulsatile pulmonary blood flow beneficial to patients with bidirectional Glenn? J Thorac Cardiovasc Surg 2013;145:451–4. 12. Xie B, Zhang JF, Shetty DP. Bidirectional Glenn shunt: 170 cases. Asian Cardiovasc Thorac Ann 2001;9:196–9. 13. Zhang WM, Zhang J, Wu ZK, Bai J. Abudushalamu. Bidirectional Glenn procedure for the treatment of complex cyanotic congenital heart diseases. Chin J Clin Thorac Cardiovasc Surg 2014;21:482–5. 14. Pankaj G, Kona SM, Kotturathu MC. Bidirectional cavopulmonary shunt with accessory pulmonary blood flow. Asian Cardiovasc Thorac Ann 2001;9:279–81. 15. Yuji N, Takeshi H. Long-term results of modified Fontan operation for single ventricle patients associated with atrioventricular valve regurgitation. Ann Thorac Surg 2013;96: 211–8. 16. Schrurer MA, Hill EG, Vasuki N, et al. Survival after bidirectional cavopulmonary anastomosis: analysis of preoperative risk factors. J Thorac Cardiovasc Surg 2007;134:82–9. 17. Schreiber C, Cleuziou J, Cornelsen JK, Horer J, Eicken A, Lange R. Bidirectional cavopulmonary connection without additional pulmonary blood flow as an ideal staging for functional univentricular hearts. Eur J Cardiothorac Surg 2008;34:550–5. 18. Yoshida M, Yamaguchi M, Yoshimura N, Murakami H, Matsuhisa H, Okita Y. Appropriate additional pulmonary blood flow at the bidirectional Glenn procedure is useful for completion of total cavopulmonary connection. Ann Thorac Surg 2005;80:976–81. 19. Vettukattil JJ, Slavik Z, Lamb RK, et al. Intrapulmonary arteriovenous shunting may be universal phenomenon in patients with the superior cavopulmonary anastomosis: a radionuclide study. Heart 2000;83:425–8. 20. Kurotobi S, Sano T, Kogaki S, et al. Bidirectional cavopulmonary shunt with right ventricular outflow patency: the impact of pulsatility on pulmonary endothelial function. J Thorac Cardiovasc Surg 2001;121:1161–8. 21. Tanoue Y, Kado H, Boku N, et al. Three hundred and thirtythree experiences with the bidirectional Glenn procedure in a single institute. Interact Cardiovasc Thorac Surg 2007;6: 97–101. 22. Rogers LS, Glatz AC, Ravishankar C, et al. 18 years of the Fontan operation at single institution. J Am Coll Cardiol 2012;60:1018–25. 23. Tan AM, Iyengar AJ, Donath S, et al. Fontan completion rate and outcomes after bidirectional cavopulmonary shunt. Eur J Cardiothorac Surg 2010;38:59–65. 24. Ganigara M, Parbhu A, Varghese R, Pavithran S, Valliatu J, Nair R. Extracardiac Fontan operation after late
Ann Thorac Surg 2016;-:-–-
bidirectional Glenn shunt. Asian Cardiovasc Thorac Ann 2010;18:253–9. 25. Suchaya S, Yupada P. Factors influencing survival in patients after bidirectional Glenn shunt. Asian Cardiovasc Thorac Ann 2008;16:381–6. 26. Sun Q, Wan D, Liu J, et al. Influence of antegrade pulmonary blood flow on the hemodynamic performance of bidirectional cavopulmonary anastomosis: a numerical study. Med Eng Phys 2009;31:227–33. 27. Lodge AJ, Rjchik J, Nicolson SC, Ittenbach RF, Spary TL, Gaynor JW. Improving outcomes in functional single
ZHANG ET AL OUTCOMES OF BDG IN CHINA
9
ventricle and total anomalous pulmonary venous connection. Ann Thorac Surg 2004;78:1688–95. 28. Lee TM, Aiyagari R, Hirsch JC, Ohye RG, Bove EL, Devaney EJ. Risk factor analysis for second-stage palliation of single ventricle anatomy. Ann Thorac Surg 2012;93:614–9. 29. Khairy P, Poirier N, Mercier LA. Univentricular heart. Circulation 2007;115:800–12. 30. Hu Z, Yuan X, Rao K, Zheng Z, Hu S. National trend in congenital heart disease mortality in China during 2003 to 2010: a population-based study. J Thorac Cardiovasc Surg 2014;148:596–602.