- Email: [email protected]
Systemic to pulmonary venous collaterals in adults with single ventricle physiology after cavopulmonary palliation
International Journal of Cardiology 189 (2015) 159–163
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Systemic to pulmonary venous collaterals in adults with single ventricle physiology after cavopulmonary palliation Gentian Lluri a,⁎, Daniel S. Levi b, Jamil Aboulhosn a a b
Ahmanson/UCLA Adult Congenital Heart Disease Center, Division of Cardiology, United States UCLA Division of Pediatric Cardiology, United States
a r t i c l e
i n f o
Article history: Received 6 December 2014 Received in revised form 12 March 2015 Accepted 9 April 2015 Available online 10 April 2015 Keywords: Fontan Venous collaterals Cyanosis Adult congenital heart disease Single ventricle physiology
a b s t r a c t Objectives: To assess the frequency, anatomic characteristics, and associations of systemic to pulmonary venous collaterals in adult patients undergoing cardiac catheterization after a Fontan operation. Additionally, the embryologic basis for the presence of venous collaterals is reviewed. Methods: Cardiac catheterization data was reviewed for 66 adults with single ventricle physiology and a Fontan palliation. Results: There were a total of 66 patients that underwent catheterization between 2004 and 2014 at the Ahmanson/UCLA Adult Congenital Heart Disease Center. There were 24 males and 42 females. Systemic venous to pulmonary venous collaterals were present in 38 patients (58%), most commonly originating from the right brachiocephalic vein (35%), azygous vein (20%) and superior vena cava (13%). Trans-catheter interventional closure was performed in 27/38 (71%) of patients with venous collaterals. At baseline these patients had lower oxygen saturation when compared to those not requiring intervention, 85.6% ± 6.1% vs 89.9% ± 5.4%, p b 0.05. At 6 months, the ambulatory systemic saturation improved from 85.6% ± 6.1% to 91.8% ± 6.4%, p b 0.05. At two years follow-up, the ambulatory systemic saturation had decreased to 90.5% ± 4.1% (p b 0.05). Conclusion: In adults with single ventricle physiology and prior Fontan surgery, systemic venous collaterals are common and can be percutaneously occluded at minimal risk with resultant improvement in systemic oxygen saturation on long term follow up. When evaluated from a developmental standpoint, 85% of collaterals are above the diaphragm and could be secondary to recanalization of the collateral veins, an embryological connection between systemic and pulmonary veins. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction The development of the systemic and pulmonary venous systems is inextricably associated with the cardiac development [1]. By four weeks in utero, the human fetal venous system is composed of three symmetric pairs of veins. These pairs include the vitelline veins originating from the yolk sac, the umbilical veins originating from the chorion, and the cardinal veins originating from the embryo all draining into the sinus venosus (Fig. 1A). The liver buds develop simultaneously and the developing sinusoids of the liver interrupt the vitelline veins dividing them into two segments: a distal segment from the liver to the yolk sac (which eventually becomes the portal vein) and a proximal segment from the liver to the heart (Fig. 1B–D). The distal left umbilical vein is the major venous connection from the placenta. Subsequently, the left umbilical vein anastomoses with the portal vein and ductus venosus to drain into the inferior vena cava (Fig. 1B–E). The anterior and ⁎ Corresponding author at: Ahmanson/UCLA Adult Congenital Heart Disease Center, Division of Cardiology, David Geffen School of Medicine at UCLA, 650 Charles Young Drive, A2-237 CHS, Los Angeles, CA 90095, United States. E-mail address: [email protected] (G. Lluri).
http://dx.doi.org/10.1016/j.ijcard.2015.04.065 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
posterior cardinal veins drain the bulk of the venous system, respectively, from the cranial and caudal sections of the body into a common cardinal vein, which drains into the sinus venosus of the developing heart. During the same time, the common pulmonary vein canalizes as a midline structure and it connects the developing pulmonary venous plexus to the common atrium (Fig. 1C) [1,2]. The presence of four pulmonary orifices does not occur until later in development, around 8 weeks (Fig. 1D). Simultaneously, venous collaterals connect the developing systemic veins to the pulmonary veins and it is the regression of these collaterals that leaves the pulmonary blood return connected only to the left atrium (Fig. 1D–E) [1,2]. Patients born with single ventricle physiology often undergo surgical cavo-pulmonary palliation, which relies on the flow of the systemic venous blood directly to the pulmonary vascular bed without the support of a pumping chamber. The superior cavopulmonary shunt (Glenn) is often used as an intermediate step prior to the Fontan operation which results in complete cavopulmonary connection [3–5], or less commonly as a final palliation in those that are not suitable candidates for the Fontan operation [6]. In patients with Glenn palliation, the superior vena cava (SVC) is directly anastamosed to the pulmonary artery (PA) resulting in elevated pressure in the SVC and PA. However,
160
G. Lluri et al. / International Journal of Cardiology 189 (2015) 159–163
Fig. 1. Development of the systemic and pulmonary venous systems and the regression of the pulmonary plexus. A) In the embryo, at four weeks there are three paired veins: the vitelline veins, umbilical veins, and cardinal veins. B and C). At five to six weeks, the liver grows and interrupts vitelline and umbilical veins. At six weeks, the common pulmonary vein drains into the left side of the common atrium (C). At around eight weeks, four pulmonary veins can be seen and venous collaterals connecting the pulmonary veins to the systemic veins can be appreciated (D). By birth, the venous collaterals have fully regressed (E).
the inferior vena cava (IVC) demonstrates equal pressure with the right atrium (RA) and the left atrium (LA), assuming that the atria have a nonrestrictive communication. As a result, the SVC pressure (equal to PA pressure) is higher than IVC pressure (equal to LA pressure) to the degree of transpulmonary gradient. Subsequently, it is hypothesized that this pressure gradient leads to development of collateral venous connections between the higher pressure SVC and low pressure IVC or pulmonary veins, serving as a “pop-off” from a high pressure to a low pressure circuit [7,8]. In addition to this well-established “pathway,” many patients will develop numerous collateral venous channels after complete cavopulmonary palliation with the Fontan operation where both the SVC and IVC return is directed to the pulmonary artery [9–12]. These venous collaterals allow deoxygenated systemic venous blood to “bypass” the lungs and therefore result in progressive cyanosis and decreased functional capacity. However, they may play an important role in limiting the pressure elevation within the cavopulmonary circuit. Several prior studies have characterized the presence of systemic to pulmonary venous collaterals in infants and children but not in adults [12–14]. This study seeks to characterize the anatomic and clinical features of venous collaterals in adult patients.
angiographic findings before and following intervention. All patients were followed in ACHDC outpatient clinic within 6 months of the procedure. Intermediate and long-term saturation data were also gathered.
2.2. Surgical data Prior details of cardiac surgical interventions were collected for all patients, including type of cavopulmonary connection, age at the time of surgery and the interval from the operation to the cardiac catheterization.
Table 2 Anatomical and surgical data. Anatomical data
Total = 66
2. Methods
Tricuspid atresia (%) DILV (%) DORV (%) Hypoplastic left heart syndrome (%) Pulmonary atresia (%) Hypoplastic right ventricle (%) Unbalanced atrioventricular canal (%)
38% 24% 18% 11% 5% 3% 1%
2.1. Patient selection
Surgical data
Total = 66
RA to PA Fontan (%) Lateral Fontan (%) Modified RA to RVOT Fontan (%) Extracardiac Fontan (%)
20% 38% 12% 30%
Between July 2004 and May 2014, cardiac catheterization studies performed at the Ahmanson/UCLA Adult Congenital Heart Disease Center (ACHDC) on adults with single ventricle physiology who had undergone prior Fontan palliation were reviewed. Only patients that had complete angiographic assessment of the Fontan, superior and inferior systemic venous and pulmonary arterial circuits were included. Sixtysix patients were identified and records were reviewed for the type of congenital heart disease, type of surgery, age at surgery, presence of any long-term complications, and age at cardiac catheterization. Catheterization procedures were reviewed for hemodynamic and Table 1 Patient characteristics and long-term complications. Total = 66 Male (%) Mean age at time of surgery (years) Mean age at time of cardiac catheterization (years) Presence of atrial arrhythmias (%) Prior episode of thromboembolic disease (%) Protein loosing enteropathy (%) Values are shown as percentages or mean ± SD where appropriate.
36 11.5 ± 8.7 29.2 ± 8.2 61 5 1.5
Values are shown as percentages. DILV = double inlet left ventricle; DORV = double outlet right ventricle; RA = right atrium; PA = pulmonary artery; RVOT = right ventricular outflow tract.
Table 3 Angiographic data. Origin of collaterals
Total = 40
Right brachiocephalic vein (%) Azygous vein (%) SVC (%) IVC (%) Thebesian veins (%) Hepatic veins (%) Left brachiocephalic vein (%) Left SVC (%) Right subclavian vein (%)
35% 20% 13% 10% 8% 5% 5% 2% 2%
Values are shown as percentages. SVC = superior vena cava; IVC = inferior vena cava.
G. Lluri et al. / International Journal of Cardiology 189 (2015) 159–163
2.3. Cineangiogram and hemodynamic analysis Cardiac catheterizations were performed using standard techniques, under general anesthesia. All baseline and follow-up saturation measurements were made on room air (21% FIO2). The location and origin of systemic-to-pulmonary venous collaterals, defined as vessels with blood flow from the Fontan circuit to the pulmonary veins or “pulmonary venous” atrium. The size of the collateral was estimated at its origin and characterized as large (if equal or greater than 4 mm) and small (if smaller than 4 mm). When closure of these systemic venous collateral vessels was performed, the type of closure
161
device was recorded and the success of embolization was evaluated angiographically.
2.4. Statistical analysis Data was collected retrospectively from patient records. Data were reported as mean +/− standard deviation and Prism 6 program (San Diego, CA) was used for statistical calculations. The paired t test was used to compare the means of continuous variable. Differences were considered significant when the p value was less than 0.05.
Fig. 2. Angiograms of systemic to pulmonary venous collaterals (anterioposterior view). A) Venous collateral from the innominate vein to the left and right pulmonary veins in a 34 yearold patient with a lateral tunnel Fontan. B) Venous collateral from the azygous vein to the right pulmonary veins in a 28 year-old male with tricuspid atresia and lateral tunnel Fontan operation. C) Venous collateral from SVC to right pulmonary veins in a 25 year-old female. D) Venous collateral from IVC to the left pulmonary veins in a 24 year-old female with unbalanced AV canal and a lateral tunnel Fontan. E) Venous collateral from the right upper hepatic vein to the right pulmonary veins in a 25 year-old female with hypoplastic left ventricle and extra-cardiac Fontan. F) A network of Thebesian veins (arrow) draining from the Fontan into the right atrium in a 24 year-old female with double inlet left ventricle and lateral tunnel Fontan.
162
G. Lluri et al. / International Journal of Cardiology 189 (2015) 159–163
3. Results
changes in Fontan pressures immediately following collateral occlusion (14.7 ± 2.7 mm Hg vs 15.0 ± 2.6 mm Hg, p = NS). There were no procedural complications of trans-catheter collateral closure reported (Table 4).
3.1. Patient characteristics There were a total of 66 adult patients who were analyzed (Table 1), 24 (36%) males and 42 (64%) females. The mean age at most recent surgery was 11.5 ± 8.7 years old. The mean age at cardiac catheterization was 29.2 ± 8.2 years old. Long-term complications post Fontan surgery were present in 53 patients (80%), atrial arrhythmia was the most common long-term complication, present in 40 (61%). In addition, 3 (5%) had a prior thromboembolic event and 1 (1.5%) patient had a diagnosis of protein losing enteropathy. At the time of data analysis, 13 patients (20%) had no evidence of long-term complications. 3.2. Anatomical and surgical data Of the 66 patients studied (Table 2), 25 (38%) had tricuspid atresia, 16 (24%) had double inlet left ventricle, 12 (18%) had double outlet right ventricle, 7 (11%) had hypoplastic left heart syndrome, 3 (5%) had pulmonary atresia (all three patients with pulmonary atresia also have hypoplastic right ventricle), 2 (3%) had hypoplastic right ventricle, and 1 (1%) patient had unbalanced atrioventricular canal. With regard to the type of Fontan surgery, 13 (20%) patients had RA to PA type Fontan, 25 (38%) patients had lateral Fontan, 8 (12%) patients had modified RA t o RVOT Fontan, 20 (30%) patients had extracardiac Fontan. 3.3. Angiographic and hemodynamic data There were 38 (58%) patients with a total of 40 venous collaterals identified. Out of these 40 collaterals (Table 3), the origin of the venous collaterals was from the right brachiocephalic vein in 14 (35%) (Fig. 2A), azygous vein in 8 (20%) patients (Fig. 2B), SVC in 5 (13%) (Fig. 2C), IVC in 4 (10%) (Fig. 2D), hepatic veins in 2 (5%) (Fig. 2E), Thebesian veins in 3 (8%) (Fig. 2F, arrow), left SVC in 2 (5%), left brachiocephalic in 1 (2%), and right subclavian in 1 (2%) of the patients. When evaluated from a developmental and anatomical standpoint, there were 34 (85%) systemic to pulmonary vein collaterals above the diaphragm and 6 (15%) below the diaphragm. Out of 38 patients with these collaterals, intervention was performed in 27 (71%) of these patients (41% of total patients). There was no difference in the mean age of patients who were intervened on (29 ± 8.3 years old) and on the patients with no collaterals or no intervention (28 ± 8.5 years old), p = 0.4. Furthermore, the Fontan circuit pressures were no different among the two groups, 14.7 ± 2.7 mm Hg in patients that underwent intervention and 15.1 ± 2.4 mm Hg in patients with no intervention, p = 0.64. However, prior to any intervention, the oxygen saturations were significantly lower in the patients that underwent intervention (85.6% ± 6.1%) when compared to patients with no intervention (89.9% ± 5.4%), p b 0.05. Six months post intervention; the systemic oxygen saturation increased from 85.6% ± 6.1% to 91.8% ± 6.4% (p b 0.05). Out of 27 patients who underwent intervention, 16 (59%) patients had long-term follow-up (defined as ≥2 years) with a mean oxygen saturation declining to 90.5 ± 4.1% (p b 0.05) at two years. There were no significant
4. Discussion Systemic to pulmonary venous collaterals are common findings in patients after Fontan surgery, however their etiology is still debated [14,12,13]. Interestingly during development, there are venous collaterals that connect the primitive pulmonary veins with the systemic veins [1,2]. Eventually, as the embryo develops these venous collaterals regress. The vast majority of the systemic to pulmonary venous/left atrial collaterals seen in Fontan patients are supradiaphragmatic and likely represent embryologic structures that “recanalize” due to elevated systemic venous pressures in patients without two-ventricle physiology. However, de novo angiogenesis, although unlikely, cannot be completely ruled out as a contributing factor in the formation of venous collateral vessels. Interestingly, this study did not find a difference in systemic venous pressure in patients with or without venous collaterals. Also, age at time of surgery or age at time of cardiac catheterization did not differ in patients with or without venous collaterals. Patients, who underwent trans-catheter venous collaterals closure, had significantly lower oxygen saturations at baseline. The saturations did not immediately improve at the time of the procedure but were noted to improve on intermediate and long-term follow-up. The systemic oxygen saturation declined between 6 months and 2 years of follow-up post intervention, however, the saturation remained significantly higher than it was at baseline (pre-intervention). The Fontan pressures did not increase acutely with embolization of collaterals. Of note, patient age, Fontan type and systemic venous pressure did not predict the presence of venous collaterals. Therefore, the only predictor of venous collateral presence in this study was decreased systemic oxygen saturation. This study is the first and largest review of angiographic and hemodynamic data in adults with single ventricle physiology and Fontan palliation, including intermediate and long-term clinical follow-up. Systemic to pulmonary venous collaterals are commonly found in this population but their presence is not predicted by duration of Fontan, systemic venous pressure or patient age. Interventional closure of veno-venous collaterals in adults is associated with intermediate and long-term improvements in systemic oxygen saturation with a low risk of complications.
Conflicts of interest statement No conflicts of interest to declare. Funding None.
Table 4 Hemodynamic data.
Age (years) Fontan pressure (mm Hg). Pre/post Oxygen saturation (%). Baseline Oxygen saturation (%). 6 months Oxygen saturation (%). 24 months Values are shown as mean ± SD. ⁎ Compared to baseline. ⁎⁎ Compared to 6 months.
Intervention (N = 27)
No intervention (N = 39)
p value
29 ± 8.3 14.7 ± 2.7/15.0 ± 2.6 85.6 ± 6.1 91.8 ± 6.4 90.5 ± 4.1
28 ± 8.5 15.1 ± 2.4 89.9 ± 5.4
0.4 0.64 b0.05 b0.05⁎ b0.05⁎, b0.05⁎⁎
G. Lluri et al. / International Journal of Cardiology 189 (2015) 159–163
References [1] G. van den Berg, A.F. Moorman, Development of the pulmonary vein and the systemic venous sinus: an interactive 3D overview, PLoS One 6 (2011) e22055. [2] S. Webb, M. Kanani, R.H. Anderson, M.K. Richardson, N.A. Brown, Development of the human pulmonary vein and its incorporation in the morphologically left atrium, Cardiol. Young 11 (2001) 632–642. [3] R.A. Hopkins, B.E. Armstrong, G.A. Serwer, R.J. Peterson, H.N. Oldham Jr., Physiological rationale for a bidirectional cavopulmonary shunt. A versatile complement to the Fontan principle, J. Thorac. Cardiovasc. Surg. 90 (1985) 391–398. [4] N.D. Bridges, R.A. Jonas, J.E. Mayer, M.F. Flanagan, J.F. Keane, A.R. Castaneda, Bidirectional cavopulmonary anastomosis as interim palliation for high-risk Fontan candidates. Early results, Circulation 82 (1990) IV170–IV176. [5] A.K. Pridjian, A.M. Mendelsohn, F.M. Lupinetti, R.H. Beekman III, M. Dick II, G. Serwer, et al., Usefulness of the bidirectional Glenn procedure as staged reconstruction for the functional single ventricle, Am. J. Cardiol. 71 (1993) 959–962. [6] V.M. Reddy, J.R. Liddicoat, F.L. Hanley, Primary bidirectional superior cavopulmonary shunt in infants between 1 and 4 months of age, Ann. Thorac. Surg. 59 (1995) 1120–1125 (discussion 5–6). [7] M.A. Gatzoulis, E.A. Shinebourne, A.N. Redington, M.L. Rigby, S.Y. Ho, D.F. Shore, Increasing cyanosis early after cavopulmonary connection caused by abnormal systemic venous channels, Br. Heart J. 73 (1995) 182–186.
163
[8] L.H. Filippini, C. Ovaert, D.G. Nykanen, R.M. Freedom, Reopening of persistent left superior caval vein after bidirectional cavopulmonary connections, Heart 79 (1998) 509–512. [9] S. Clapp, W.R. Morrow, Development of superior vena cava to pulmonary vein fistulae following modified Fontan operation: case report of a rare anomaly and embolization therapy, Pediatr. Cardiol. 19 (1998) 363–365. [10] P. Fernandez-Martorell, M.S. Sklansky, V.W. Lucas, I.A. Kashani, M.W. Cocalis, S.W. Jamieson, et al., Accessory hepatic vein to pulmonary venous atrium as a cause of cyanosis after the Fontan operation, Am. J. Cardiol. 77 (1996) 1386–1387. [11] M. Heinemann, J. Breuer, V. Steger, E. Steil, L. Sieverding, G. Ziemer, Incidence and impact of systemic venous collateral development after Glenn and Fontan procedures, Thorac. Cardiovasc. Surg. 49 (2001) 172–178. [12] D.B. McElhinney, V.M. Reddy, F.L. Hanley, P. Moore, Systemic venous collateral channels causing desaturation after bidirectional cavopulmonary anastomosis: evaluation and management, J. Am. Coll. Cardiol. 30 (1997) 817–824. [13] A.G. Magee, B.W. McCrindle, J. Mawson, L.N. Benson, W.G. Williams, R.M. Freedom, Systemic venous collateral development after the bidirectional cavopulmonary anastomosis. Prevalence and predictors, J. Am. Coll. Cardiol. 32 (1998) 502–508. [14] H. Sugiyama, S.J. Yoo, W. Williams, L.N. Benson, Characterization and treatment of systemic venous to pulmonary venous collaterals seen after the Fontan operation, Cardiol. Young 13 (2003) 424–430.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Systemic to pulmonary venous collaterals in adults with single ventricle physiology after cavopulmonary palliation Gentian Lluri a,⁎, Daniel S. Levi b, Jamil Aboulhosn a a b
Ahmanson/UCLA Adult Congenital Heart Disease Center, Division of Cardiology, United States UCLA Division of Pediatric Cardiology, United States
a r t i c l e
i n f o
Article history: Received 6 December 2014 Received in revised form 12 March 2015 Accepted 9 April 2015 Available online 10 April 2015 Keywords: Fontan Venous collaterals Cyanosis Adult congenital heart disease Single ventricle physiology
a b s t r a c t Objectives: To assess the frequency, anatomic characteristics, and associations of systemic to pulmonary venous collaterals in adult patients undergoing cardiac catheterization after a Fontan operation. Additionally, the embryologic basis for the presence of venous collaterals is reviewed. Methods: Cardiac catheterization data was reviewed for 66 adults with single ventricle physiology and a Fontan palliation. Results: There were a total of 66 patients that underwent catheterization between 2004 and 2014 at the Ahmanson/UCLA Adult Congenital Heart Disease Center. There were 24 males and 42 females. Systemic venous to pulmonary venous collaterals were present in 38 patients (58%), most commonly originating from the right brachiocephalic vein (35%), azygous vein (20%) and superior vena cava (13%). Trans-catheter interventional closure was performed in 27/38 (71%) of patients with venous collaterals. At baseline these patients had lower oxygen saturation when compared to those not requiring intervention, 85.6% ± 6.1% vs 89.9% ± 5.4%, p b 0.05. At 6 months, the ambulatory systemic saturation improved from 85.6% ± 6.1% to 91.8% ± 6.4%, p b 0.05. At two years follow-up, the ambulatory systemic saturation had decreased to 90.5% ± 4.1% (p b 0.05). Conclusion: In adults with single ventricle physiology and prior Fontan surgery, systemic venous collaterals are common and can be percutaneously occluded at minimal risk with resultant improvement in systemic oxygen saturation on long term follow up. When evaluated from a developmental standpoint, 85% of collaterals are above the diaphragm and could be secondary to recanalization of the collateral veins, an embryological connection between systemic and pulmonary veins. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction The development of the systemic and pulmonary venous systems is inextricably associated with the cardiac development [1]. By four weeks in utero, the human fetal venous system is composed of three symmetric pairs of veins. These pairs include the vitelline veins originating from the yolk sac, the umbilical veins originating from the chorion, and the cardinal veins originating from the embryo all draining into the sinus venosus (Fig. 1A). The liver buds develop simultaneously and the developing sinusoids of the liver interrupt the vitelline veins dividing them into two segments: a distal segment from the liver to the yolk sac (which eventually becomes the portal vein) and a proximal segment from the liver to the heart (Fig. 1B–D). The distal left umbilical vein is the major venous connection from the placenta. Subsequently, the left umbilical vein anastomoses with the portal vein and ductus venosus to drain into the inferior vena cava (Fig. 1B–E). The anterior and ⁎ Corresponding author at: Ahmanson/UCLA Adult Congenital Heart Disease Center, Division of Cardiology, David Geffen School of Medicine at UCLA, 650 Charles Young Drive, A2-237 CHS, Los Angeles, CA 90095, United States. E-mail address: [email protected] (G. Lluri).
http://dx.doi.org/10.1016/j.ijcard.2015.04.065 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
posterior cardinal veins drain the bulk of the venous system, respectively, from the cranial and caudal sections of the body into a common cardinal vein, which drains into the sinus venosus of the developing heart. During the same time, the common pulmonary vein canalizes as a midline structure and it connects the developing pulmonary venous plexus to the common atrium (Fig. 1C) [1,2]. The presence of four pulmonary orifices does not occur until later in development, around 8 weeks (Fig. 1D). Simultaneously, venous collaterals connect the developing systemic veins to the pulmonary veins and it is the regression of these collaterals that leaves the pulmonary blood return connected only to the left atrium (Fig. 1D–E) [1,2]. Patients born with single ventricle physiology often undergo surgical cavo-pulmonary palliation, which relies on the flow of the systemic venous blood directly to the pulmonary vascular bed without the support of a pumping chamber. The superior cavopulmonary shunt (Glenn) is often used as an intermediate step prior to the Fontan operation which results in complete cavopulmonary connection [3–5], or less commonly as a final palliation in those that are not suitable candidates for the Fontan operation [6]. In patients with Glenn palliation, the superior vena cava (SVC) is directly anastamosed to the pulmonary artery (PA) resulting in elevated pressure in the SVC and PA. However,
160
G. Lluri et al. / International Journal of Cardiology 189 (2015) 159–163
Fig. 1. Development of the systemic and pulmonary venous systems and the regression of the pulmonary plexus. A) In the embryo, at four weeks there are three paired veins: the vitelline veins, umbilical veins, and cardinal veins. B and C). At five to six weeks, the liver grows and interrupts vitelline and umbilical veins. At six weeks, the common pulmonary vein drains into the left side of the common atrium (C). At around eight weeks, four pulmonary veins can be seen and venous collaterals connecting the pulmonary veins to the systemic veins can be appreciated (D). By birth, the venous collaterals have fully regressed (E).
the inferior vena cava (IVC) demonstrates equal pressure with the right atrium (RA) and the left atrium (LA), assuming that the atria have a nonrestrictive communication. As a result, the SVC pressure (equal to PA pressure) is higher than IVC pressure (equal to LA pressure) to the degree of transpulmonary gradient. Subsequently, it is hypothesized that this pressure gradient leads to development of collateral venous connections between the higher pressure SVC and low pressure IVC or pulmonary veins, serving as a “pop-off” from a high pressure to a low pressure circuit [7,8]. In addition to this well-established “pathway,” many patients will develop numerous collateral venous channels after complete cavopulmonary palliation with the Fontan operation where both the SVC and IVC return is directed to the pulmonary artery [9–12]. These venous collaterals allow deoxygenated systemic venous blood to “bypass” the lungs and therefore result in progressive cyanosis and decreased functional capacity. However, they may play an important role in limiting the pressure elevation within the cavopulmonary circuit. Several prior studies have characterized the presence of systemic to pulmonary venous collaterals in infants and children but not in adults [12–14]. This study seeks to characterize the anatomic and clinical features of venous collaterals in adult patients.
angiographic findings before and following intervention. All patients were followed in ACHDC outpatient clinic within 6 months of the procedure. Intermediate and long-term saturation data were also gathered.
2.2. Surgical data Prior details of cardiac surgical interventions were collected for all patients, including type of cavopulmonary connection, age at the time of surgery and the interval from the operation to the cardiac catheterization.
Table 2 Anatomical and surgical data. Anatomical data
Total = 66
2. Methods
Tricuspid atresia (%) DILV (%) DORV (%) Hypoplastic left heart syndrome (%) Pulmonary atresia (%) Hypoplastic right ventricle (%) Unbalanced atrioventricular canal (%)
38% 24% 18% 11% 5% 3% 1%
2.1. Patient selection
Surgical data
Total = 66
RA to PA Fontan (%) Lateral Fontan (%) Modified RA to RVOT Fontan (%) Extracardiac Fontan (%)
20% 38% 12% 30%
Between July 2004 and May 2014, cardiac catheterization studies performed at the Ahmanson/UCLA Adult Congenital Heart Disease Center (ACHDC) on adults with single ventricle physiology who had undergone prior Fontan palliation were reviewed. Only patients that had complete angiographic assessment of the Fontan, superior and inferior systemic venous and pulmonary arterial circuits were included. Sixtysix patients were identified and records were reviewed for the type of congenital heart disease, type of surgery, age at surgery, presence of any long-term complications, and age at cardiac catheterization. Catheterization procedures were reviewed for hemodynamic and Table 1 Patient characteristics and long-term complications. Total = 66 Male (%) Mean age at time of surgery (years) Mean age at time of cardiac catheterization (years) Presence of atrial arrhythmias (%) Prior episode of thromboembolic disease (%) Protein loosing enteropathy (%) Values are shown as percentages or mean ± SD where appropriate.
36 11.5 ± 8.7 29.2 ± 8.2 61 5 1.5
Values are shown as percentages. DILV = double inlet left ventricle; DORV = double outlet right ventricle; RA = right atrium; PA = pulmonary artery; RVOT = right ventricular outflow tract.
Table 3 Angiographic data. Origin of collaterals
Total = 40
Right brachiocephalic vein (%) Azygous vein (%) SVC (%) IVC (%) Thebesian veins (%) Hepatic veins (%) Left brachiocephalic vein (%) Left SVC (%) Right subclavian vein (%)
35% 20% 13% 10% 8% 5% 5% 2% 2%
Values are shown as percentages. SVC = superior vena cava; IVC = inferior vena cava.
G. Lluri et al. / International Journal of Cardiology 189 (2015) 159–163
2.3. Cineangiogram and hemodynamic analysis Cardiac catheterizations were performed using standard techniques, under general anesthesia. All baseline and follow-up saturation measurements were made on room air (21% FIO2). The location and origin of systemic-to-pulmonary venous collaterals, defined as vessels with blood flow from the Fontan circuit to the pulmonary veins or “pulmonary venous” atrium. The size of the collateral was estimated at its origin and characterized as large (if equal or greater than 4 mm) and small (if smaller than 4 mm). When closure of these systemic venous collateral vessels was performed, the type of closure
161
device was recorded and the success of embolization was evaluated angiographically.
2.4. Statistical analysis Data was collected retrospectively from patient records. Data were reported as mean +/− standard deviation and Prism 6 program (San Diego, CA) was used for statistical calculations. The paired t test was used to compare the means of continuous variable. Differences were considered significant when the p value was less than 0.05.
Fig. 2. Angiograms of systemic to pulmonary venous collaterals (anterioposterior view). A) Venous collateral from the innominate vein to the left and right pulmonary veins in a 34 yearold patient with a lateral tunnel Fontan. B) Venous collateral from the azygous vein to the right pulmonary veins in a 28 year-old male with tricuspid atresia and lateral tunnel Fontan operation. C) Venous collateral from SVC to right pulmonary veins in a 25 year-old female. D) Venous collateral from IVC to the left pulmonary veins in a 24 year-old female with unbalanced AV canal and a lateral tunnel Fontan. E) Venous collateral from the right upper hepatic vein to the right pulmonary veins in a 25 year-old female with hypoplastic left ventricle and extra-cardiac Fontan. F) A network of Thebesian veins (arrow) draining from the Fontan into the right atrium in a 24 year-old female with double inlet left ventricle and lateral tunnel Fontan.
162
G. Lluri et al. / International Journal of Cardiology 189 (2015) 159–163
3. Results
changes in Fontan pressures immediately following collateral occlusion (14.7 ± 2.7 mm Hg vs 15.0 ± 2.6 mm Hg, p = NS). There were no procedural complications of trans-catheter collateral closure reported (Table 4).
3.1. Patient characteristics There were a total of 66 adult patients who were analyzed (Table 1), 24 (36%) males and 42 (64%) females. The mean age at most recent surgery was 11.5 ± 8.7 years old. The mean age at cardiac catheterization was 29.2 ± 8.2 years old. Long-term complications post Fontan surgery were present in 53 patients (80%), atrial arrhythmia was the most common long-term complication, present in 40 (61%). In addition, 3 (5%) had a prior thromboembolic event and 1 (1.5%) patient had a diagnosis of protein losing enteropathy. At the time of data analysis, 13 patients (20%) had no evidence of long-term complications. 3.2. Anatomical and surgical data Of the 66 patients studied (Table 2), 25 (38%) had tricuspid atresia, 16 (24%) had double inlet left ventricle, 12 (18%) had double outlet right ventricle, 7 (11%) had hypoplastic left heart syndrome, 3 (5%) had pulmonary atresia (all three patients with pulmonary atresia also have hypoplastic right ventricle), 2 (3%) had hypoplastic right ventricle, and 1 (1%) patient had unbalanced atrioventricular canal. With regard to the type of Fontan surgery, 13 (20%) patients had RA to PA type Fontan, 25 (38%) patients had lateral Fontan, 8 (12%) patients had modified RA t o RVOT Fontan, 20 (30%) patients had extracardiac Fontan. 3.3. Angiographic and hemodynamic data There were 38 (58%) patients with a total of 40 venous collaterals identified. Out of these 40 collaterals (Table 3), the origin of the venous collaterals was from the right brachiocephalic vein in 14 (35%) (Fig. 2A), azygous vein in 8 (20%) patients (Fig. 2B), SVC in 5 (13%) (Fig. 2C), IVC in 4 (10%) (Fig. 2D), hepatic veins in 2 (5%) (Fig. 2E), Thebesian veins in 3 (8%) (Fig. 2F, arrow), left SVC in 2 (5%), left brachiocephalic in 1 (2%), and right subclavian in 1 (2%) of the patients. When evaluated from a developmental and anatomical standpoint, there were 34 (85%) systemic to pulmonary vein collaterals above the diaphragm and 6 (15%) below the diaphragm. Out of 38 patients with these collaterals, intervention was performed in 27 (71%) of these patients (41% of total patients). There was no difference in the mean age of patients who were intervened on (29 ± 8.3 years old) and on the patients with no collaterals or no intervention (28 ± 8.5 years old), p = 0.4. Furthermore, the Fontan circuit pressures were no different among the two groups, 14.7 ± 2.7 mm Hg in patients that underwent intervention and 15.1 ± 2.4 mm Hg in patients with no intervention, p = 0.64. However, prior to any intervention, the oxygen saturations were significantly lower in the patients that underwent intervention (85.6% ± 6.1%) when compared to patients with no intervention (89.9% ± 5.4%), p b 0.05. Six months post intervention; the systemic oxygen saturation increased from 85.6% ± 6.1% to 91.8% ± 6.4% (p b 0.05). Out of 27 patients who underwent intervention, 16 (59%) patients had long-term follow-up (defined as ≥2 years) with a mean oxygen saturation declining to 90.5 ± 4.1% (p b 0.05) at two years. There were no significant
4. Discussion Systemic to pulmonary venous collaterals are common findings in patients after Fontan surgery, however their etiology is still debated [14,12,13]. Interestingly during development, there are venous collaterals that connect the primitive pulmonary veins with the systemic veins [1,2]. Eventually, as the embryo develops these venous collaterals regress. The vast majority of the systemic to pulmonary venous/left atrial collaterals seen in Fontan patients are supradiaphragmatic and likely represent embryologic structures that “recanalize” due to elevated systemic venous pressures in patients without two-ventricle physiology. However, de novo angiogenesis, although unlikely, cannot be completely ruled out as a contributing factor in the formation of venous collateral vessels. Interestingly, this study did not find a difference in systemic venous pressure in patients with or without venous collaterals. Also, age at time of surgery or age at time of cardiac catheterization did not differ in patients with or without venous collaterals. Patients, who underwent trans-catheter venous collaterals closure, had significantly lower oxygen saturations at baseline. The saturations did not immediately improve at the time of the procedure but were noted to improve on intermediate and long-term follow-up. The systemic oxygen saturation declined between 6 months and 2 years of follow-up post intervention, however, the saturation remained significantly higher than it was at baseline (pre-intervention). The Fontan pressures did not increase acutely with embolization of collaterals. Of note, patient age, Fontan type and systemic venous pressure did not predict the presence of venous collaterals. Therefore, the only predictor of venous collateral presence in this study was decreased systemic oxygen saturation. This study is the first and largest review of angiographic and hemodynamic data in adults with single ventricle physiology and Fontan palliation, including intermediate and long-term clinical follow-up. Systemic to pulmonary venous collaterals are commonly found in this population but their presence is not predicted by duration of Fontan, systemic venous pressure or patient age. Interventional closure of veno-venous collaterals in adults is associated with intermediate and long-term improvements in systemic oxygen saturation with a low risk of complications.
Conflicts of interest statement No conflicts of interest to declare. Funding None.
Table 4 Hemodynamic data.
Age (years) Fontan pressure (mm Hg). Pre/post Oxygen saturation (%). Baseline Oxygen saturation (%). 6 months Oxygen saturation (%). 24 months Values are shown as mean ± SD. ⁎ Compared to baseline. ⁎⁎ Compared to 6 months.
Intervention (N = 27)
No intervention (N = 39)
p value
29 ± 8.3 14.7 ± 2.7/15.0 ± 2.6 85.6 ± 6.1 91.8 ± 6.4 90.5 ± 4.1
28 ± 8.5 15.1 ± 2.4 89.9 ± 5.4
0.4 0.64 b0.05 b0.05⁎ b0.05⁎, b0.05⁎⁎
G. Lluri et al. / International Journal of Cardiology 189 (2015) 159–163
References [1] G. van den Berg, A.F. Moorman, Development of the pulmonary vein and the systemic venous sinus: an interactive 3D overview, PLoS One 6 (2011) e22055. [2] S. Webb, M. Kanani, R.H. Anderson, M.K. Richardson, N.A. Brown, Development of the human pulmonary vein and its incorporation in the morphologically left atrium, Cardiol. Young 11 (2001) 632–642. [3] R.A. Hopkins, B.E. Armstrong, G.A. Serwer, R.J. Peterson, H.N. Oldham Jr., Physiological rationale for a bidirectional cavopulmonary shunt. A versatile complement to the Fontan principle, J. Thorac. Cardiovasc. Surg. 90 (1985) 391–398. [4] N.D. Bridges, R.A. Jonas, J.E. Mayer, M.F. Flanagan, J.F. Keane, A.R. Castaneda, Bidirectional cavopulmonary anastomosis as interim palliation for high-risk Fontan candidates. Early results, Circulation 82 (1990) IV170–IV176. [5] A.K. Pridjian, A.M. Mendelsohn, F.M. Lupinetti, R.H. Beekman III, M. Dick II, G. Serwer, et al., Usefulness of the bidirectional Glenn procedure as staged reconstruction for the functional single ventricle, Am. J. Cardiol. 71 (1993) 959–962. [6] V.M. Reddy, J.R. Liddicoat, F.L. Hanley, Primary bidirectional superior cavopulmonary shunt in infants between 1 and 4 months of age, Ann. Thorac. Surg. 59 (1995) 1120–1125 (discussion 5–6). [7] M.A. Gatzoulis, E.A. Shinebourne, A.N. Redington, M.L. Rigby, S.Y. Ho, D.F. Shore, Increasing cyanosis early after cavopulmonary connection caused by abnormal systemic venous channels, Br. Heart J. 73 (1995) 182–186.
163
[8] L.H. Filippini, C. Ovaert, D.G. Nykanen, R.M. Freedom, Reopening of persistent left superior caval vein after bidirectional cavopulmonary connections, Heart 79 (1998) 509–512. [9] S. Clapp, W.R. Morrow, Development of superior vena cava to pulmonary vein fistulae following modified Fontan operation: case report of a rare anomaly and embolization therapy, Pediatr. Cardiol. 19 (1998) 363–365. [10] P. Fernandez-Martorell, M.S. Sklansky, V.W. Lucas, I.A. Kashani, M.W. Cocalis, S.W. Jamieson, et al., Accessory hepatic vein to pulmonary venous atrium as a cause of cyanosis after the Fontan operation, Am. J. Cardiol. 77 (1996) 1386–1387. [11] M. Heinemann, J. Breuer, V. Steger, E. Steil, L. Sieverding, G. Ziemer, Incidence and impact of systemic venous collateral development after Glenn and Fontan procedures, Thorac. Cardiovasc. Surg. 49 (2001) 172–178. [12] D.B. McElhinney, V.M. Reddy, F.L. Hanley, P. Moore, Systemic venous collateral channels causing desaturation after bidirectional cavopulmonary anastomosis: evaluation and management, J. Am. Coll. Cardiol. 30 (1997) 817–824. [13] A.G. Magee, B.W. McCrindle, J. Mawson, L.N. Benson, W.G. Williams, R.M. Freedom, Systemic venous collateral development after the bidirectional cavopulmonary anastomosis. Prevalence and predictors, J. Am. Coll. Cardiol. 32 (1998) 502–508. [14] H. Sugiyama, S.J. Yoo, W. Williams, L.N. Benson, Characterization and treatment of systemic venous to pulmonary venous collaterals seen after the Fontan operation, Cardiol. Young 13 (2003) 424–430.