bs_bs_banner
doi:10.1111/jgh.12583
H E PAT O L O G Y
Portopulmonary venous anastomosis in balloon-occluded retrograde transvenous obliteration for the treatment of gastric varices Shuji Kariya,* Atsushi Komemushi,* Miyuki Nakatani,* Rie Yoshida,* Yumiko Kono,* Tomokuni Shiraishi† and Noboru Tanigawa* *Department of Radiology, Kansai Medical University, Hirakata, and †Department of Radiology, Ishikiriseiki Hospital, Higashiosaka, Osaka, Japan
Key words carbon dioxide, esophageal and gastric varices, hypertension, microbubbles, portal. Accepted for publication 26 February 2014. Correspondence Dr Shuji Kariya, Department of Radiology, Kansai Medical University, 2-5-1 Shinmachi, Hirakata, Osaka 573-1010, Japan. Email: [email protected]
Abstract Background and Aim: Several reports have described portopulmonary venous anastomosis (PPVA). However, in balloon-occluded retrograde transvenous obliteration (BRTO), attention has not been paid to paradoxical embolism. The objective of this study was to investigate the existence of a right–left shunt due to PPVA when the drainage vein is occluded by a balloon during BRTO. Methods: The subjects were 19 patients who underwent BRTO. Whether PPVA was present was confirmed on balloon-occluded retrograde transvenous venography (BRTV). After BRTV, a retrograde bolus injection of 20 mL of carbon dioxide (CO2) via the balloon catheter was performed under balloon occlusion, and the flow of bubbles into both ventricles was observed with four-chamber view echocardiography. During the same balloon occlusion, bolus injection of CO2 into the inferior vena cava was performed, followed by echocardiography. Results: PPVA was confirmed on BRTV in four patients (21.1%). On echocardiography with retrograde CO2 injection, bubbles were confirmed in the left ventricle in six patients (31.6%). On echocardiography with CO2 injection into the inferior vena cava, bubbles were not confirmed in the left ventricle in any cases. Conclusions: When the draining vein was occluded with a balloon and blood flow in a gastrorenal or gastrocaval shunt was stopped during BRTO, PPVA was confirmed in 21.1% of cases on retrograde angiography, and a right–left shunt was confirmed in 31.6% of cases on echocardiography.
Introduction In 1996, Kanagawa et al. first reported the effectiveness of balloon-occluded retrograde transvenous obliteration (BRTO) for treating gastric varices, and several reports have since described the benefits of this approach.1–14 Previous studies have described the use of ethanolamine oleate as the embolic material for portosystemic shunts. Occlusion of the collateral channels, which affects the success or failure of BRTO, has been carried out by stepwise injection of ethanolamine oleate or embolization with a coil, ethanol, or 50% glucose solution.6–14 Recently, particle-like materials, foamed polidocanol with air, N-butyl-cyanoacrylate, and other materials have been used for BRTO.15–19 However, due to the presence of a portopulmonary venous anastomosis (PPVA) in the collateral channels, we were concerned that paradoxical embolism might occur if these were used for BRTO.20–31 We have, therefore, been screening for PPVA prior to embolization and selecting embolic materials and methods depending on its presence. 1522
The objective of this study was to investigate the presence and frequency of right–left shunt due to PPVA when the drainage vein is occluded by a balloon during BRTO treatment.
Materials and methods Study design and subjects. The study protocols for this retrospective analysis were approved by our institutional review board. The requirement to obtain informed consent was waived. Data were gathered retrospectively by reviewing clinical records, including images. Twenty consecutive patients underwent 21 BRTO procedures for gastric varices between January 2010 and March 2012. In one patient, the first BRTO was clinically unsuccessful because a residual gastric varix was confirmed, necessitating a second BRTO that proved successful. Balloon occlusion was impossible in one patient, and surgical treatment was subsequently performed. The gastric varix was successfully visualized by balloon-occluded
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
S Kariya et al.
retrograde transvenous venography (BRTV) in 19 patients. Thus, the subjects were 19 patients (12 men, 7 women) who underwent 20 BRTO procedures. The mean age at the time of intervention was 65 ± 11 years (range, 37–78 years). Indications for BRTO were a ruptured gastric varix or gastric varices judged to be at risk of rupture on endoscopic examination. Risk of rupture was defined endoscopically as a growing varix, a medium-sized beaded varix with a red color sign, or a large tuberous or tumorous varix. In this study, the existence of PPVA was defined as the detection of a vein connected from the portosystemic shunt to the pulmonary vein on BRTV.
Procedure. An 8- or 10-Fr, J-shaped sheath-introducer (Super Sheath; Medikit, Miyazaki, Japan) was inserted into the vena cava using a right internal jugular or right femoral vein approach. From the sheath, a cobra or multipurpose-shaped occlusion balloon catheter with an open tip (Selecon MP Catheter II; Terumo Clinical Supply, Gifu, Japan or Candis; Medikit) was inserted into the draining vein (the left adrenal vein in 18 patients with a gastrorenal shunt, the left inferior phrenic vein in one patient with a gastrocaval shunt) of the gastric varix, and the balloon was inflated for BRTV. BRTV was performed manually by bolus injection of 30 mL of iodinated contrast medium. The radiation field for angiography extended from the hilar region to the renal vein. Whether PPVA was present was confirmed on BRTV. After BRTV, retrograde bolus injection of 20 mL of carbon dioxide (CO2) was performed manually via the balloon catheter under balloon occlusion, and the flow of bubbles into both ventricles was observed under four-chamber view echocardiography. After the bubbles disappeared, bolus injection of 20 mL of CO2 into the inferior vena cava was performed manually via the side arm of the sheath-introducer under balloon occlusion, and the same echocardiographic observations were made. Injection via the sheath-introducer into the systemic vein was performed to distinguish intracardiac or pulmonary circulation from right–left shunting. In cases where bubbles were observed in the left ventricle, coil embolization of the collateral channel was attempted whenever possible, or an attempt was made to advance the balloon catheter to the location where the collateral channel was not visualized on BRTV. In such cases, N-butyl-cyanoacrylate or a particle-like material was not used for embolization because of the risk of paradoxical embolism. In cases in which the portal vein was visualized and stagnation of the contrast medium in the gastric varix was seen on BRTV, 5% ethanolamine oleate–iopamidol (EOI) was injected. The injection was completed at the time the 5% EOI reached the afferent vein to the gastric varix. After 4–18 hours with the balloon occlusion in place and the 5% EOI in the gastric varix, as much of the 5% EOI as possible was aspirated through the catheter, and the procedure was concluded. Contrast-enhanced computed tomography or endoscopy was performed one month after BRTO, and thrombosis or regression of varices was confirmed.
Results In four of the 19 patients, BRTV depicted one of the left inferior pulmonary veins and confirmed the existence of PPVA (Table 1)
Portopulmonary venous anastomosis
(Figs 1–3). In all four patients, the PPVA had the afferent vein arising from the inferior phrenic vein. On echocardiography, when CO2 was injected in a retrograde manner into the shunt under balloon occlusion, bubbles were confirmed in the right ventricle in all 19 patients, while bubbles flowing into the left ventricle were confirmed in 6 (31.6%) (Table 1). In all four patients with PPVA, bubbles flowing into the left ventricle were confirmed. When CO2 was injected from the side arm of the sheath-introducer, bubbles were confirmed in the right ventricle in all 19 patients, but in no cases in the left ventricle. Thus, the possibility of a right–left shunt via a PPVA was shown in six (31.6%). In two of the four patients in whom a PPVA was confirmed on BRTV, the balloon catheter was inserted to the drainage vein passing the bifurcation of the collateral channels, and BRTO was performed. In one of the four patients, the collateral channels were embolized by the coil, and BRTO was performed. In one of the four patients, despite coil embolization of some collaterals, drainage flow from the gastric varix could not be stopped, and BRTO was unsuccessful. In all cases, the procedure was completed without any complications.
Discussion In successful BRTO, the extent to which closure can be achieved in collateral channels that could potentially become new drainage routes following occlusion of the main draining vein from the gastric varix by the balloon is important.6,9,10 Various reports have described the types of embolic material and techniques for collateral channel embolization.6–19 The collateral channels can form a PPVA via the bronchial venous system and create a right–left shunt (Fig. 4). The presence of PPVAs in veins that can serve as such collateral channels was reported by Schoenmackers et al. in 1953.20 In a summary of subsequent reports, Kumar et al. found that PPVA is present in 1–30% of cases of portal hypertension.21 We therefore considered the possibility of paradoxical embolism with embolization or injection of a sclerosing agent in these collateral channels, even with BRTO. We conjectured that, in BRTO, the possibility of a right–left shunt via a PPVA is increased because the portal vein pressure is raised even further by occlusion of the drainage veins. In light of this possibility that paradoxical embolism might also occur in BRTO, we have been screening for the presence of PPVA before embolization. If PPVA is present, we do not use N-butylcyanoacrylate, foamed polidocanol with air, or a particle-like embolic material. Sano et al. reported the characteristic findings of the inferior pulmonary vein due to PPVA as “sudden spurting of contrast medium,” “counterclockwise arch-like movement or flux of contrast medium,” and “synchronization of this motion of the flux of contrast medium with cardiac contraction” on cineportography.25,27 In the present study, findings corresponding to those of Sano et al. were confirmed in four of the 19 patients (21.1%). However, in these cases, the amount of contrast medium that reached the afferent vein to the PPVA was small, and these findings were attributed to poor imaging conditions and difficult visualization. As the inferior pulmonary vein presents characteristic dynamic features, cine-BRTV should have
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1523
1524
72 73 62 77 65 62 64 75 60 55 74 78 61 75 70 49 53 74 37 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19† 19†
F M M F F M M M M M M M M M F M F F F F
Sex
Yes Yes No No No No No No No No No No No Yes No No No Yes No No
Detection of PPVA by angiography
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Detection of bubbles in the right ventricle by echocardiography
Injection via balloon catheter
Yes Yes Yes No No No No No No No No No No Yes No Yes No Yes No No
Detection of bubbles in the left ventricle by echocardiography Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Detection of bubbles in the right ventricle by echocardiography No No No No No No No No No No No No No No No No No No No No
Detection of bubbles in the left ventricle by echocardiography
Injection via side arm of sheath-introducer
The first BRTO was clinically unsuccessful because a residual gastric varix was confirmed, and a second BRTO was therefore performed successfully. BRTO, balloon occluded retrograde transvenous obliteration; PPVA, portopulmonary venous anastomosis.
†
Age (year)
Detection of right–left shunts with the draining vein blocked by the balloon
Patient
Table 1
Success Success Success Success Success Success Success Success Success Success Success Success Success Success Success Success Success Failure Failure Success
Results of BRTO
Portopulmonary venous anastomosis S Kariya et al.
© 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527
S Kariya et al.
Figure 1 A 73-year-old man with alcoholic cirrhosis and a gastric varix. Balloon-occluded retrograde transvenous venography shows the left inferior pulmonary vein (black arrow), the afferent vein arising from the inferior phrenic vein (white arrow), and the gastric varix (asterisk). Thus, a portopulmonary venous anastomosis is confirmed.
Figure 2 A 75-year-old man with cirrhosis type C and a gastric varix. Balloon-occluded retrograde transvenous venography shows the left inferior pulmonary vein (black arrow), two afferent veins arising from the inferior phrenic vein (white arrow), and the gastric varix (asterisk). Thus, a portopulmonary venous anastomosis is confirmed.
Portopulmonary venous anastomosis
Figure 3 A 74-year-old women with cirrhosis type C and a gastric varix. Balloon-occluded retrograde transvenous venography shows the left inferior pulmonary vein (black arrow), the afferent vein arising from the inferior phrenic vein (white arrow), and the gastric varix (asterisk). Thus, a portopulmonary venous anastomosis is confirmed.
been used for identification rather than static imaging. Sano et al. also identified the inferior pulmonary vein on cine images.25,27 In the present study, PPVA was confirmed in four of 19 patients (21.1%) on BRTV, and a right–left shunt was confirmed in six of 19 patients (31.6%) on echocardiography. Thus, echocardiography is considered more sensitive. In two of the six patients in whom the bubbles flowing into the left ventricle were confirmed, PPVA was not detected. We considered two potential causes for this. The first may have been that, because the quantity of the contrast medium that reached the pulmonary vein was too small, it could not be detected on BRTV. Another was that CO2, which would have different hemodynamics from contrast medium, can more easily fill many collaterals and then may often go into the PPVA. Liquid sclerosant and particle materials may be different from CO2 in their hemodynamics, but foam sclerosing agent may be similar to CO2, and paradoxical embolization might similarly occur when using air. However, EOI is similar to contrast medium in its hemodynamics, and the quantity that reaches a PPVA is also too small, so that paradoxical embolization might not easily occur. There have been reports of the occurrence of paradoxical embolism in the treatment of gastric varices by endoscopy, transjugular intrahepatic portosystemic shunt, and percutaneous transhepatic obliteration by means of particle-like embolic material or N-butyl-cyanoacrylate.32–36 However, no case of
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1525
Portopulmonary venous anastomosis
S Kariya et al.
References
Figure 4 When blood flow in the draining vein serving as a drainage route for the varix is stopped by balloon occlusion in balloon-occluded retrograde obliteration, collateral channels often become new drainage routes. The collateral channels can include the paraesophageal, mediastinal, azygos, and bronchial veins. These veins can form a portopulmonary venous anastomosis via the bronchial venous system and create a right–left shunt. AdV, adrenal vein; AzV, azygos vein; BV, bronchial vein; IPV, inferior phrenic vein; IVC, inferior vena cava; MeV, mediastinal vein; PEV, paraesophageal vein; PoV, portal vein; PuV, pulmonary vein; RV, renal vein; SpV, splenic vein.
paradoxical embolism has been reported in BRTO. We conjectured that this is because, in most cases, embolization in BRTO has been performed using ethanolamine oleate, 50% glucose solution, ethanol, or coils. As long as the coil does not become dislodged, paradoxical embolism cannot occur, and even if small amounts of ethanolamine oleate, 50% glucose solution, or ethanol should reach the left heart, they do not have an embolic effect. However, even the small amounts of particle-like embolic material, N-butyl-cyanoacrylate, and foamed polidocanol with air that have been used in BRTO in recent years may cause paradoxical embolism. Given the possibility of PPVA, attention must be paid to migration of thrombus and air during BRTO procedures. The limitations of the present study were its retrospective design and small sample size. When balloon occlusion of the portosystemic shunt was carried out in BRTO, the presence of a right–left shunt was confirmed by echocardiography after this blood flow had been shut off, but whether this was due to PPVA remains conjecture. In conclusion, when the draining vein was occluded with a balloon in BRTO, PPVA was confirmed in 21.1% of cases on retrograde angiography, and a right–left shunt was confirmed in 31.6% of cases on echocardiography.
1526
1 Kanagawa H, Mima S, Kouyama H, Gotoh K, Uchida T, Okuda K. Treatment of gastric fundal varices by balloon-occluded retrograde transvenous obliteration. J. Gastroenterol. Hepatol. 1996; 11: 51–8. 2 Matsumoto A, Hamamoto N, Nomura T et al. Balloon-occluded retrograde transvenous obliteration of high risk gastric fundal varices. Am. J. Gastroenterol. 1999; 94: 643–9. 3 Akahane T, Iwasaki T, Kobayashi N et al. Changes in liver function parameters after occlusion of gastrorenal shunts with balloon-occluded retrograde transvenous obliteration. Am. J. Gastroenterol. 1997; 92: 1026–30. 4 Koito K, Namieno T, Nagakawa T, Morita K. Balloon-occluded retrograde transvenous obliteration for gastric varices with gastrorenal or gastrocaval collaterals. AJR Am. J. Roentgenol. 1996; 167: 1317–20. 5 Sonomura T, Sato M, Kishi K et al. Balloon-occluded retrograde transvenous obliteration for gastric varices: a feasibility study. Cardiovasc. Intervent. Radiol. 1998; 21: 27–30. 6 Hirota S, Matsumoto S, Tomita M, Sako M, Kono M. Retrograde transvenous obliteration of gastric varices. Radiology 1999; 211: 349–56. 7 Takahashi K, Yamada T, Hyodoh H et al. Selective balloon-occluded retrograde sclerosis of gastric varices using a coaxial microcatheter system. AJR Am. J. Roentgenol. 2001; 177: 1091–3. 8 Fukuda T, Hirota S, Sugimura K. Long-term results of balloon-occluded retrograde transvenous obliteration for the treatment of gastric varices and hepatic encephalopathy. J. Vasc. Interv. Radiol. 2001; 12: 327–36. 9 Kiyosue H, Mori H, Matsumoto S, Yamada Y, Hori Y, Okino Y. Transcatheter obliteration of gastric varices. Part 1. Anatomic classification. Radiographics 2003; 23: 911–20. 10 Kiyosue H, Mori H, Matsumoto S, Yamada Y, Hori Y, Okino Y. Transcatheter obliteration of gastric varices: Part 2. Strategy and techniques based on hemodynamic features. Radiographics 2003; 23: 921–37; discussion 37. 11 Nishida N, Ninoi T, Kitayama T et al. Dual balloon-occluded retrograde transvenous obliteration of gastric varix draining into the left adrenal vein and left inferior phrenic vein. Cardiovasc. Intervent. Radiol. 2004; 27: 560–2. 12 Minamiguchi H, Kawai N, Sato M et al. Dual microcatheter retrograde transvenous obliteration of gastric varices: coil embolization as a substitute for balloon occlusion. Case Rep. Gastroenterol. 2012; 6: 74–81. 13 Yamagami T, Tanaka O, Yoshimatsu R, Miura H, Nishimura T. Value of embolisation of collateral veins from gastric varices before balloon-occluded retrograde transvenous obliteration. J. Med. Imaging Radiat. Oncol. 2011; 55: 26–32. 14 Yamagami T, Kato T, Hirota T, Yoshimatsu R, Matsumoto T, Nishimura T. Infusion of 50% glucose solution before injection of ethanolamine oleate during balloon-occluded retrograde transvenous obliteration. Australas. Radiol. 2007; 51: 334–8. 15 Choi SY, Won JY, Kim KA, Lee do Y, Lee KH. Foam sclerotherapy using polidocanol for balloon-occluded retrograde transvenous obliteration (BRTO). Eur. Radiol. 2011; 21: 122–9. 16 Sabri SS, Swee W, Turba UC et al. Bleeding gastric varices obliteration with balloon-occluded retrograde transvenous obliteration using sodium tetradecyl sulfate foam. J. Vasc. Interv. Radiol. 2011; 22: 309–16. 17 Teo TK, Sabri SS, Turba UC, Saad WE, Angle JF. Obliteration of bleeding peristomal varices with balloon-occluded retrograde transvenous obliteration using sodium tetradecyl sulfate foam. J. Vasc. Interv. Radiol. 2011; 22: 1049–51.
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
S Kariya et al.
18 Koizumi J, Hashimoto T, Myojin K et al. Balloon-occluded retrograde transvenous obliteration of gastric varices: use of CT-guided foam sclerotherapy to optimize technique. AJR Am. J. Roentgenol. 2012; 199: 200–7. 19 Hashimoto R, Sofue K, Takeuchi Y, Shibamoto K, Arai Y. Successful balloon-occluded retrograde transvenous obliteration for bleeding duodenal varices using cyanoacrylate. World J. Gastroenterol. 2013; 19: 951–4. 20 Schoenmackers J, Vieten H. [Portocaval and portopulmonary anastomoses in postmortem portography]. Fortschr. Geb. Rontgenstr. 1953; 79: 488–98. 21 Kumar A, Gonzalez G, Wilkinson L et al. Computed tomography findings of spontaneous porto-pulmonary shunts in 3 patients with portal hypertension. J. Thorac. Imaging 2010; 25: W70–4. 22 Doehner GA, Ruzicka FF, Jr, Rousselot LM, Hoffman G. The portal venous system: on its pathological roentgen anatomy. Radiology 1956; 66: 206–17. 23 Calabresi P, Abelmann WH. Porto-caval and porto-pulmonary anastomoses in Laennec’s cirrhosis and in heart failure. J. Clin. Invest. 1957; 36: 1257–65. 24 Shaldon S, Caesar J, Chiandussi L. The detection of porta-pulmonary shunting in portal cirrhosis using radio-active krypton (Kr-85). Minerva Cardioangiol. 1961; 9: 441–3. 25 Sano A, Kuroda Y, Moriyasu F et al. Porto-pulmonary venous anastomosis in portal hypertension demonstrated by percutaneous transhepatic cine-portography. Radiology 1982; 144: 479–84. 26 Nunez D, Russell E, Yrizarry J, Pereiras R, Viamonte M, Jr. Portosystemic communications studied by transhepatic portography. Radiology 1978; 127: 75–9. 27 Sano A, Nishizawa S, Sasai K et al. Contrast echocardiography in detection of portopulmonary venous anastomosis. AJR Am. J. Roentgenol. 1984; 142: 137–40.
Portopulmonary venous anastomosis
28 Blackburn CR. Acquired portal-pulmonary venous anastomosis complicating partial oesophago-gastrectomy in a patient with portal hypertension. Thorax 1956; 11: 30–5. 29 Massumi RA, Rios JC, Ticktin HE. Hemodynamic abnormalities and venous admixture in portal cirrhosis. Am. J. Med. Sci. 1965; 250: 275–83. 30 Miura H, Yamagami T, Tanaka O, Yoshimatsu R. Portopulmonary venous anastomosis detected at balloon-occluded retrograde transvenous obliteration for gastric varices. J. Vasc. Interv. Radiol. 2013; 24: 131–4. 31 Yamagami T, Yoshimatsu R, Miura H, Yamada K, Minami M. Successful balloon-occluded retrograde transvenous obliteration of gastric varix via pericardiacophrenic vein after embolization of portopulmonary venous anastomosis. J. Vasc. Interv. Radiol. 2013; 24: 137–9. 32 Ponec RJ, Kowdley KV. Paradoxical cerebral emboli after transjugular intrahepatic portosystemic shunt and coil embolization for treatment of duodenal varices. Am. J. Gastroenterol. 1997; 92: 1372–3. 33 Ellman BA, Curry TS 3rd, Glotzbach RE, Simpson PR. Systemic embolization as a complication of transhepatic venography. Radiology 1981; 141: 67–71. 34 Mendez G, Jr, Russell E. Gastrointestinal varices: percutaneous transheptic therapeutic embolization in 54 patients. AJR Am. J. Roentgenol. 1980; 135: 1045–50. 35 S Saracco G, Giordanino C, Roberto N et al. Fatal multiple systemic embolisms after injection of cyanoacrylate in bleeding gastric varices of a patient who was noncirrhotic but with idiopathic portal hypertension. Gastrointest. Endosc. 2007; 65: 345–7. 36 Tan YM, Goh KL, Kamarulzaman A et al. Multiple systemic embolisms with septicemia after gastric variceal obliteration with cyanoacrylate. Gastrointest. Endosc. 2002; 55: 276–8.
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1527
doi:10.1111/jgh.12583
H E PAT O L O G Y
Portopulmonary venous anastomosis in balloon-occluded retrograde transvenous obliteration for the treatment of gastric varices Shuji Kariya,* Atsushi Komemushi,* Miyuki Nakatani,* Rie Yoshida,* Yumiko Kono,* Tomokuni Shiraishi† and Noboru Tanigawa* *Department of Radiology, Kansai Medical University, Hirakata, and †Department of Radiology, Ishikiriseiki Hospital, Higashiosaka, Osaka, Japan
Key words carbon dioxide, esophageal and gastric varices, hypertension, microbubbles, portal. Accepted for publication 26 February 2014. Correspondence Dr Shuji Kariya, Department of Radiology, Kansai Medical University, 2-5-1 Shinmachi, Hirakata, Osaka 573-1010, Japan. Email: [email protected]
Abstract Background and Aim: Several reports have described portopulmonary venous anastomosis (PPVA). However, in balloon-occluded retrograde transvenous obliteration (BRTO), attention has not been paid to paradoxical embolism. The objective of this study was to investigate the existence of a right–left shunt due to PPVA when the drainage vein is occluded by a balloon during BRTO. Methods: The subjects were 19 patients who underwent BRTO. Whether PPVA was present was confirmed on balloon-occluded retrograde transvenous venography (BRTV). After BRTV, a retrograde bolus injection of 20 mL of carbon dioxide (CO2) via the balloon catheter was performed under balloon occlusion, and the flow of bubbles into both ventricles was observed with four-chamber view echocardiography. During the same balloon occlusion, bolus injection of CO2 into the inferior vena cava was performed, followed by echocardiography. Results: PPVA was confirmed on BRTV in four patients (21.1%). On echocardiography with retrograde CO2 injection, bubbles were confirmed in the left ventricle in six patients (31.6%). On echocardiography with CO2 injection into the inferior vena cava, bubbles were not confirmed in the left ventricle in any cases. Conclusions: When the draining vein was occluded with a balloon and blood flow in a gastrorenal or gastrocaval shunt was stopped during BRTO, PPVA was confirmed in 21.1% of cases on retrograde angiography, and a right–left shunt was confirmed in 31.6% of cases on echocardiography.
Introduction In 1996, Kanagawa et al. first reported the effectiveness of balloon-occluded retrograde transvenous obliteration (BRTO) for treating gastric varices, and several reports have since described the benefits of this approach.1–14 Previous studies have described the use of ethanolamine oleate as the embolic material for portosystemic shunts. Occlusion of the collateral channels, which affects the success or failure of BRTO, has been carried out by stepwise injection of ethanolamine oleate or embolization with a coil, ethanol, or 50% glucose solution.6–14 Recently, particle-like materials, foamed polidocanol with air, N-butyl-cyanoacrylate, and other materials have been used for BRTO.15–19 However, due to the presence of a portopulmonary venous anastomosis (PPVA) in the collateral channels, we were concerned that paradoxical embolism might occur if these were used for BRTO.20–31 We have, therefore, been screening for PPVA prior to embolization and selecting embolic materials and methods depending on its presence. 1522
The objective of this study was to investigate the presence and frequency of right–left shunt due to PPVA when the drainage vein is occluded by a balloon during BRTO treatment.
Materials and methods Study design and subjects. The study protocols for this retrospective analysis were approved by our institutional review board. The requirement to obtain informed consent was waived. Data were gathered retrospectively by reviewing clinical records, including images. Twenty consecutive patients underwent 21 BRTO procedures for gastric varices between January 2010 and March 2012. In one patient, the first BRTO was clinically unsuccessful because a residual gastric varix was confirmed, necessitating a second BRTO that proved successful. Balloon occlusion was impossible in one patient, and surgical treatment was subsequently performed. The gastric varix was successfully visualized by balloon-occluded
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
S Kariya et al.
retrograde transvenous venography (BRTV) in 19 patients. Thus, the subjects were 19 patients (12 men, 7 women) who underwent 20 BRTO procedures. The mean age at the time of intervention was 65 ± 11 years (range, 37–78 years). Indications for BRTO were a ruptured gastric varix or gastric varices judged to be at risk of rupture on endoscopic examination. Risk of rupture was defined endoscopically as a growing varix, a medium-sized beaded varix with a red color sign, or a large tuberous or tumorous varix. In this study, the existence of PPVA was defined as the detection of a vein connected from the portosystemic shunt to the pulmonary vein on BRTV.
Procedure. An 8- or 10-Fr, J-shaped sheath-introducer (Super Sheath; Medikit, Miyazaki, Japan) was inserted into the vena cava using a right internal jugular or right femoral vein approach. From the sheath, a cobra or multipurpose-shaped occlusion balloon catheter with an open tip (Selecon MP Catheter II; Terumo Clinical Supply, Gifu, Japan or Candis; Medikit) was inserted into the draining vein (the left adrenal vein in 18 patients with a gastrorenal shunt, the left inferior phrenic vein in one patient with a gastrocaval shunt) of the gastric varix, and the balloon was inflated for BRTV. BRTV was performed manually by bolus injection of 30 mL of iodinated contrast medium. The radiation field for angiography extended from the hilar region to the renal vein. Whether PPVA was present was confirmed on BRTV. After BRTV, retrograde bolus injection of 20 mL of carbon dioxide (CO2) was performed manually via the balloon catheter under balloon occlusion, and the flow of bubbles into both ventricles was observed under four-chamber view echocardiography. After the bubbles disappeared, bolus injection of 20 mL of CO2 into the inferior vena cava was performed manually via the side arm of the sheath-introducer under balloon occlusion, and the same echocardiographic observations were made. Injection via the sheath-introducer into the systemic vein was performed to distinguish intracardiac or pulmonary circulation from right–left shunting. In cases where bubbles were observed in the left ventricle, coil embolization of the collateral channel was attempted whenever possible, or an attempt was made to advance the balloon catheter to the location where the collateral channel was not visualized on BRTV. In such cases, N-butyl-cyanoacrylate or a particle-like material was not used for embolization because of the risk of paradoxical embolism. In cases in which the portal vein was visualized and stagnation of the contrast medium in the gastric varix was seen on BRTV, 5% ethanolamine oleate–iopamidol (EOI) was injected. The injection was completed at the time the 5% EOI reached the afferent vein to the gastric varix. After 4–18 hours with the balloon occlusion in place and the 5% EOI in the gastric varix, as much of the 5% EOI as possible was aspirated through the catheter, and the procedure was concluded. Contrast-enhanced computed tomography or endoscopy was performed one month after BRTO, and thrombosis or regression of varices was confirmed.
Results In four of the 19 patients, BRTV depicted one of the left inferior pulmonary veins and confirmed the existence of PPVA (Table 1)
Portopulmonary venous anastomosis
(Figs 1–3). In all four patients, the PPVA had the afferent vein arising from the inferior phrenic vein. On echocardiography, when CO2 was injected in a retrograde manner into the shunt under balloon occlusion, bubbles were confirmed in the right ventricle in all 19 patients, while bubbles flowing into the left ventricle were confirmed in 6 (31.6%) (Table 1). In all four patients with PPVA, bubbles flowing into the left ventricle were confirmed. When CO2 was injected from the side arm of the sheath-introducer, bubbles were confirmed in the right ventricle in all 19 patients, but in no cases in the left ventricle. Thus, the possibility of a right–left shunt via a PPVA was shown in six (31.6%). In two of the four patients in whom a PPVA was confirmed on BRTV, the balloon catheter was inserted to the drainage vein passing the bifurcation of the collateral channels, and BRTO was performed. In one of the four patients, the collateral channels were embolized by the coil, and BRTO was performed. In one of the four patients, despite coil embolization of some collaterals, drainage flow from the gastric varix could not be stopped, and BRTO was unsuccessful. In all cases, the procedure was completed without any complications.
Discussion In successful BRTO, the extent to which closure can be achieved in collateral channels that could potentially become new drainage routes following occlusion of the main draining vein from the gastric varix by the balloon is important.6,9,10 Various reports have described the types of embolic material and techniques for collateral channel embolization.6–19 The collateral channels can form a PPVA via the bronchial venous system and create a right–left shunt (Fig. 4). The presence of PPVAs in veins that can serve as such collateral channels was reported by Schoenmackers et al. in 1953.20 In a summary of subsequent reports, Kumar et al. found that PPVA is present in 1–30% of cases of portal hypertension.21 We therefore considered the possibility of paradoxical embolism with embolization or injection of a sclerosing agent in these collateral channels, even with BRTO. We conjectured that, in BRTO, the possibility of a right–left shunt via a PPVA is increased because the portal vein pressure is raised even further by occlusion of the drainage veins. In light of this possibility that paradoxical embolism might also occur in BRTO, we have been screening for the presence of PPVA before embolization. If PPVA is present, we do not use N-butylcyanoacrylate, foamed polidocanol with air, or a particle-like embolic material. Sano et al. reported the characteristic findings of the inferior pulmonary vein due to PPVA as “sudden spurting of contrast medium,” “counterclockwise arch-like movement or flux of contrast medium,” and “synchronization of this motion of the flux of contrast medium with cardiac contraction” on cineportography.25,27 In the present study, findings corresponding to those of Sano et al. were confirmed in four of the 19 patients (21.1%). However, in these cases, the amount of contrast medium that reached the afferent vein to the PPVA was small, and these findings were attributed to poor imaging conditions and difficult visualization. As the inferior pulmonary vein presents characteristic dynamic features, cine-BRTV should have
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1523
1524
72 73 62 77 65 62 64 75 60 55 74 78 61 75 70 49 53 74 37 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19† 19†
F M M F F M M M M M M M M M F M F F F F
Sex
Yes Yes No No No No No No No No No No No Yes No No No Yes No No
Detection of PPVA by angiography
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Detection of bubbles in the right ventricle by echocardiography
Injection via balloon catheter
Yes Yes Yes No No No No No No No No No No Yes No Yes No Yes No No
Detection of bubbles in the left ventricle by echocardiography Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Detection of bubbles in the right ventricle by echocardiography No No No No No No No No No No No No No No No No No No No No
Detection of bubbles in the left ventricle by echocardiography
Injection via side arm of sheath-introducer
The first BRTO was clinically unsuccessful because a residual gastric varix was confirmed, and a second BRTO was therefore performed successfully. BRTO, balloon occluded retrograde transvenous obliteration; PPVA, portopulmonary venous anastomosis.
†
Age (year)
Detection of right–left shunts with the draining vein blocked by the balloon
Patient
Table 1
Success Success Success Success Success Success Success Success Success Success Success Success Success Success Success Success Success Failure Failure Success
Results of BRTO
Portopulmonary venous anastomosis S Kariya et al.
© 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527
S Kariya et al.
Figure 1 A 73-year-old man with alcoholic cirrhosis and a gastric varix. Balloon-occluded retrograde transvenous venography shows the left inferior pulmonary vein (black arrow), the afferent vein arising from the inferior phrenic vein (white arrow), and the gastric varix (asterisk). Thus, a portopulmonary venous anastomosis is confirmed.
Figure 2 A 75-year-old man with cirrhosis type C and a gastric varix. Balloon-occluded retrograde transvenous venography shows the left inferior pulmonary vein (black arrow), two afferent veins arising from the inferior phrenic vein (white arrow), and the gastric varix (asterisk). Thus, a portopulmonary venous anastomosis is confirmed.
Portopulmonary venous anastomosis
Figure 3 A 74-year-old women with cirrhosis type C and a gastric varix. Balloon-occluded retrograde transvenous venography shows the left inferior pulmonary vein (black arrow), the afferent vein arising from the inferior phrenic vein (white arrow), and the gastric varix (asterisk). Thus, a portopulmonary venous anastomosis is confirmed.
been used for identification rather than static imaging. Sano et al. also identified the inferior pulmonary vein on cine images.25,27 In the present study, PPVA was confirmed in four of 19 patients (21.1%) on BRTV, and a right–left shunt was confirmed in six of 19 patients (31.6%) on echocardiography. Thus, echocardiography is considered more sensitive. In two of the six patients in whom the bubbles flowing into the left ventricle were confirmed, PPVA was not detected. We considered two potential causes for this. The first may have been that, because the quantity of the contrast medium that reached the pulmonary vein was too small, it could not be detected on BRTV. Another was that CO2, which would have different hemodynamics from contrast medium, can more easily fill many collaterals and then may often go into the PPVA. Liquid sclerosant and particle materials may be different from CO2 in their hemodynamics, but foam sclerosing agent may be similar to CO2, and paradoxical embolization might similarly occur when using air. However, EOI is similar to contrast medium in its hemodynamics, and the quantity that reaches a PPVA is also too small, so that paradoxical embolization might not easily occur. There have been reports of the occurrence of paradoxical embolism in the treatment of gastric varices by endoscopy, transjugular intrahepatic portosystemic shunt, and percutaneous transhepatic obliteration by means of particle-like embolic material or N-butyl-cyanoacrylate.32–36 However, no case of
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1525
Portopulmonary venous anastomosis
S Kariya et al.
References
Figure 4 When blood flow in the draining vein serving as a drainage route for the varix is stopped by balloon occlusion in balloon-occluded retrograde obliteration, collateral channels often become new drainage routes. The collateral channels can include the paraesophageal, mediastinal, azygos, and bronchial veins. These veins can form a portopulmonary venous anastomosis via the bronchial venous system and create a right–left shunt. AdV, adrenal vein; AzV, azygos vein; BV, bronchial vein; IPV, inferior phrenic vein; IVC, inferior vena cava; MeV, mediastinal vein; PEV, paraesophageal vein; PoV, portal vein; PuV, pulmonary vein; RV, renal vein; SpV, splenic vein.
paradoxical embolism has been reported in BRTO. We conjectured that this is because, in most cases, embolization in BRTO has been performed using ethanolamine oleate, 50% glucose solution, ethanol, or coils. As long as the coil does not become dislodged, paradoxical embolism cannot occur, and even if small amounts of ethanolamine oleate, 50% glucose solution, or ethanol should reach the left heart, they do not have an embolic effect. However, even the small amounts of particle-like embolic material, N-butyl-cyanoacrylate, and foamed polidocanol with air that have been used in BRTO in recent years may cause paradoxical embolism. Given the possibility of PPVA, attention must be paid to migration of thrombus and air during BRTO procedures. The limitations of the present study were its retrospective design and small sample size. When balloon occlusion of the portosystemic shunt was carried out in BRTO, the presence of a right–left shunt was confirmed by echocardiography after this blood flow had been shut off, but whether this was due to PPVA remains conjecture. In conclusion, when the draining vein was occluded with a balloon in BRTO, PPVA was confirmed in 21.1% of cases on retrograde angiography, and a right–left shunt was confirmed in 31.6% of cases on echocardiography.
1526
1 Kanagawa H, Mima S, Kouyama H, Gotoh K, Uchida T, Okuda K. Treatment of gastric fundal varices by balloon-occluded retrograde transvenous obliteration. J. Gastroenterol. Hepatol. 1996; 11: 51–8. 2 Matsumoto A, Hamamoto N, Nomura T et al. Balloon-occluded retrograde transvenous obliteration of high risk gastric fundal varices. Am. J. Gastroenterol. 1999; 94: 643–9. 3 Akahane T, Iwasaki T, Kobayashi N et al. Changes in liver function parameters after occlusion of gastrorenal shunts with balloon-occluded retrograde transvenous obliteration. Am. J. Gastroenterol. 1997; 92: 1026–30. 4 Koito K, Namieno T, Nagakawa T, Morita K. Balloon-occluded retrograde transvenous obliteration for gastric varices with gastrorenal or gastrocaval collaterals. AJR Am. J. Roentgenol. 1996; 167: 1317–20. 5 Sonomura T, Sato M, Kishi K et al. Balloon-occluded retrograde transvenous obliteration for gastric varices: a feasibility study. Cardiovasc. Intervent. Radiol. 1998; 21: 27–30. 6 Hirota S, Matsumoto S, Tomita M, Sako M, Kono M. Retrograde transvenous obliteration of gastric varices. Radiology 1999; 211: 349–56. 7 Takahashi K, Yamada T, Hyodoh H et al. Selective balloon-occluded retrograde sclerosis of gastric varices using a coaxial microcatheter system. AJR Am. J. Roentgenol. 2001; 177: 1091–3. 8 Fukuda T, Hirota S, Sugimura K. Long-term results of balloon-occluded retrograde transvenous obliteration for the treatment of gastric varices and hepatic encephalopathy. J. Vasc. Interv. Radiol. 2001; 12: 327–36. 9 Kiyosue H, Mori H, Matsumoto S, Yamada Y, Hori Y, Okino Y. Transcatheter obliteration of gastric varices. Part 1. Anatomic classification. Radiographics 2003; 23: 911–20. 10 Kiyosue H, Mori H, Matsumoto S, Yamada Y, Hori Y, Okino Y. Transcatheter obliteration of gastric varices: Part 2. Strategy and techniques based on hemodynamic features. Radiographics 2003; 23: 921–37; discussion 37. 11 Nishida N, Ninoi T, Kitayama T et al. Dual balloon-occluded retrograde transvenous obliteration of gastric varix draining into the left adrenal vein and left inferior phrenic vein. Cardiovasc. Intervent. Radiol. 2004; 27: 560–2. 12 Minamiguchi H, Kawai N, Sato M et al. Dual microcatheter retrograde transvenous obliteration of gastric varices: coil embolization as a substitute for balloon occlusion. Case Rep. Gastroenterol. 2012; 6: 74–81. 13 Yamagami T, Tanaka O, Yoshimatsu R, Miura H, Nishimura T. Value of embolisation of collateral veins from gastric varices before balloon-occluded retrograde transvenous obliteration. J. Med. Imaging Radiat. Oncol. 2011; 55: 26–32. 14 Yamagami T, Kato T, Hirota T, Yoshimatsu R, Matsumoto T, Nishimura T. Infusion of 50% glucose solution before injection of ethanolamine oleate during balloon-occluded retrograde transvenous obliteration. Australas. Radiol. 2007; 51: 334–8. 15 Choi SY, Won JY, Kim KA, Lee do Y, Lee KH. Foam sclerotherapy using polidocanol for balloon-occluded retrograde transvenous obliteration (BRTO). Eur. Radiol. 2011; 21: 122–9. 16 Sabri SS, Swee W, Turba UC et al. Bleeding gastric varices obliteration with balloon-occluded retrograde transvenous obliteration using sodium tetradecyl sulfate foam. J. Vasc. Interv. Radiol. 2011; 22: 309–16. 17 Teo TK, Sabri SS, Turba UC, Saad WE, Angle JF. Obliteration of bleeding peristomal varices with balloon-occluded retrograde transvenous obliteration using sodium tetradecyl sulfate foam. J. Vasc. Interv. Radiol. 2011; 22: 1049–51.
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
S Kariya et al.
18 Koizumi J, Hashimoto T, Myojin K et al. Balloon-occluded retrograde transvenous obliteration of gastric varices: use of CT-guided foam sclerotherapy to optimize technique. AJR Am. J. Roentgenol. 2012; 199: 200–7. 19 Hashimoto R, Sofue K, Takeuchi Y, Shibamoto K, Arai Y. Successful balloon-occluded retrograde transvenous obliteration for bleeding duodenal varices using cyanoacrylate. World J. Gastroenterol. 2013; 19: 951–4. 20 Schoenmackers J, Vieten H. [Portocaval and portopulmonary anastomoses in postmortem portography]. Fortschr. Geb. Rontgenstr. 1953; 79: 488–98. 21 Kumar A, Gonzalez G, Wilkinson L et al. Computed tomography findings of spontaneous porto-pulmonary shunts in 3 patients with portal hypertension. J. Thorac. Imaging 2010; 25: W70–4. 22 Doehner GA, Ruzicka FF, Jr, Rousselot LM, Hoffman G. The portal venous system: on its pathological roentgen anatomy. Radiology 1956; 66: 206–17. 23 Calabresi P, Abelmann WH. Porto-caval and porto-pulmonary anastomoses in Laennec’s cirrhosis and in heart failure. J. Clin. Invest. 1957; 36: 1257–65. 24 Shaldon S, Caesar J, Chiandussi L. The detection of porta-pulmonary shunting in portal cirrhosis using radio-active krypton (Kr-85). Minerva Cardioangiol. 1961; 9: 441–3. 25 Sano A, Kuroda Y, Moriyasu F et al. Porto-pulmonary venous anastomosis in portal hypertension demonstrated by percutaneous transhepatic cine-portography. Radiology 1982; 144: 479–84. 26 Nunez D, Russell E, Yrizarry J, Pereiras R, Viamonte M, Jr. Portosystemic communications studied by transhepatic portography. Radiology 1978; 127: 75–9. 27 Sano A, Nishizawa S, Sasai K et al. Contrast echocardiography in detection of portopulmonary venous anastomosis. AJR Am. J. Roentgenol. 1984; 142: 137–40.
Portopulmonary venous anastomosis
28 Blackburn CR. Acquired portal-pulmonary venous anastomosis complicating partial oesophago-gastrectomy in a patient with portal hypertension. Thorax 1956; 11: 30–5. 29 Massumi RA, Rios JC, Ticktin HE. Hemodynamic abnormalities and venous admixture in portal cirrhosis. Am. J. Med. Sci. 1965; 250: 275–83. 30 Miura H, Yamagami T, Tanaka O, Yoshimatsu R. Portopulmonary venous anastomosis detected at balloon-occluded retrograde transvenous obliteration for gastric varices. J. Vasc. Interv. Radiol. 2013; 24: 131–4. 31 Yamagami T, Yoshimatsu R, Miura H, Yamada K, Minami M. Successful balloon-occluded retrograde transvenous obliteration of gastric varix via pericardiacophrenic vein after embolization of portopulmonary venous anastomosis. J. Vasc. Interv. Radiol. 2013; 24: 137–9. 32 Ponec RJ, Kowdley KV. Paradoxical cerebral emboli after transjugular intrahepatic portosystemic shunt and coil embolization for treatment of duodenal varices. Am. J. Gastroenterol. 1997; 92: 1372–3. 33 Ellman BA, Curry TS 3rd, Glotzbach RE, Simpson PR. Systemic embolization as a complication of transhepatic venography. Radiology 1981; 141: 67–71. 34 Mendez G, Jr, Russell E. Gastrointestinal varices: percutaneous transheptic therapeutic embolization in 54 patients. AJR Am. J. Roentgenol. 1980; 135: 1045–50. 35 S Saracco G, Giordanino C, Roberto N et al. Fatal multiple systemic embolisms after injection of cyanoacrylate in bleeding gastric varices of a patient who was noncirrhotic but with idiopathic portal hypertension. Gastrointest. Endosc. 2007; 65: 345–7. 36 Tan YM, Goh KL, Kamarulzaman A et al. Multiple systemic embolisms with septicemia after gastric variceal obliteration with cyanoacrylate. Gastrointest. Endosc. 2002; 55: 276–8.
Journal of Gastroenterology and Hepatology 29 (2014) 1522–1527 © 2014 Journal of Gastroenterology and Hepatology Foundation and Wiley Publishing Asia Pty Ltd
1527
