Journal of Pediatric Surgery 49 (2014) 436–440
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Prehepatic portal hypertension with aneurysm of the portal vein: Unusual but treatable malformative pattern Arianna Bertocchini a,⁎, Giuseppe d’Ambrosio a, Chiara Grimaldi a, Laura del Prete a, Fabrizio di Francesco a, Piergiorgio Falappa b, Lidia Monti c, Jean de Ville de Goyet a a b c
Hepato-biliary and Transplant Surgery Unit, Department of Surgery and Transplantation Centre, Bambino Gesù Childrens Hospital, Roma, Italy Interventional Radiology Unit, Department of Surgery and Transplantation Centre, Bambino Gesù Childrens Hospital, Roma, Italy Hepato-biliary Radiology Unit, Department of Imaging, Bambino Gesù Childrens Hospital, Roma, Italy
a r t i c l e
i n f o
Article history: Received 30 March 2013 Received in revised form 9 June 2013 Accepted 10 June 2013 Key words: Intrahepatic portal vein Vein Aneurysm Rex recessus Extrahepatic portal hypertension Meso Rex bypass
a b s t r a c t Introduction: Portal vein aneurysms (PVAs) are usually located at the vein trunk or at its bifurcation, rarely intra-hepatic, or at the umbilical portion. Etiology remains unclear. Methods: Three children with PVA were identified over a 5-year period. PVA anatomy was assessed by Doppler Ultrasound, Angio CT/MRI, and trans-jugular retrograde portography. Results: Three children with intrahepatic PVA (including the umbilical portion) were identified during assessment for pre-hepatic portal hypertension: all had splenomegaly and hypersplenism. One presented with massive variceal bleeding. In two cases, a portal vein cavernoma was found, and in the third a severe stricture at the portal bifurcation was observed. Restoration of portal venous flow was achieved by a mesoRex bypass in two cases and transposing the PV into the Rex in one. High hepatopetal portal flow was restored immediately, with follow-up confirming long-term patency and resolution of signs of portal hypertension with time. Conclusions: These original observations suggest a common initial malformative pattern consisting of a portal venous stricture/web causing a post-stenotic aneurysmal dilatation of the intrahepatic portal branches complicated by thrombosis and cavernomatous transformation of the portal vein trunk. Importantly, the Meso-Rex bypass allows restoring a normal portal flow and cures the portal hypertension. © 2014 Elsevier Inc. All rights reserved.
Portal vein aneurysm (PVA) is a rare vascular malformation (0.067% incidence) [1], representing fewer than 3% of all venous aneurysms [2]. First described by Barziali and Kleckner in 1956 [3,4], almost 150 cases are now reported in literature [5] (only 68 cases in English) [6]). Morphologically, portal vein aneurysm can be fusiform or saccular [7]. Two main categories have been described: intrahepatic or extrahepatic. Extrahepatic aneurysms are more common, located at the main portal vein in 52% of cases, at the superior mesenteric– splenic vein confluence in 44%, and rarely (3% to 4%) at the level of the right or left portal branches [4]. Portal vein diameter being variable, an aneurysmal dilatation of the PV is defined according to the venous upper limit of range for age [8–10]. Intrahepatic aneurysms and those involving the umbilical portion of portal vein are rare with only few cases being reported [11–13]. Clinical presentation varies, especially in relation to the aneurysm size, the majority of patients presenting with nonspecific abdominal pain (44% to 54%), while many others are asymptomatic and diagnosis ⁎ Corresponding author. Department of Pediatric Surgery and Transplantation Center, Bambino Gesù Childrens Hospital, Piazza Sant’Onofrio 4, 00165 Roma, Italy. Tel.: +39 0668592851; +390668592425; +393383203451; fax: +390668593841. E-mail address: [email protected] (A. Bertocchini). 0022-3468/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpedsurg.2013.06.012
is incidental (25% to 38%). Rarely, patients are diagnosed with portal hypertension or gastrointestinal bleeding (7.3% to 9.8%) [3,4]. In this report, 3 children with pre-hepatic portal hypertension and intrahepatic portal aneurysms are presented, and a surgical cure is described. 1. Patients and methods Between January 2007 and December 2011, three children with PVA were identified out of a population of 105 children who were radiologically evaluated for extrahepatic portal hypertension (3%, all 3 female). Patient characteristics, clinical presentation, anatomy of the PVA, surgical correction and outcome were reviewed. Anatomical assessment of the anatomy was done by standard imaging (Doppler Ultrasound (US), Angio CT, Angio MRI) and transjugular hepatic vein catheterization (gradient measurement and retrograde portography) in all cases. 1.1. Patients 1.1.1. Patient 1 At the age of 18, she was admitted to our centre after a single episode of massive variceal bleeding a month earlier. Endoscopy had shown four F2 varices with red stripes in the lower portion of the
A. Bertocchini et al. / Journal of Pediatric Surgery 49 (2014) 436–440
esophagus, and she was treated by elastic bands positioning. She had splenomegaly (US span: 20 cm), thrombocytopenia (platelet count: 57,000/ml) and normal hepatic function tests. At radiological assessment, a portal cavernoma and aneurysm of both right and left portal branches, including a huge aneurysmal transformation of the Rex recessus, were observed. 1.1.2. Patient 2 A 5-year-old girl presented initially with variceal bleeding; esophageal varices, factor VII deficiency, splenomegaly (US span: 14.5 cm) and thrombocytopenia (platelet count: 86,000/ml). At imaging, a malformative pattern combining PVA of both portal branches extending into the Rex recessus, a normal extrahepatic portal vein trunk and a severe stricture of the main portal trunk at the level of the bifurcation were found (Figs. 1 and 2A). An attempt at percutaneous transhepatic angioplasty/stenting of the stricture in another centre failed, and she was referred to our centre for surgical treatment. 1.1.3. Patient 3 Splenomegaly and portal vein cavernoma were incidentally diagnosed in a 3-year old girl with an uneventful history and absence of previous gastrointestinal bleeding. Endoscopy showed esophageal varices F2, gastric varices and mild hypertensive gastropathy. She was in good general conditions, with splenomegaly (US span: 16 cm), thrombocytopenia (platelet count: 56,000/ml) and normal hepatic function tests. Radiological assessment showed a portal cavernoma with patent intrahepatic portal branches of a normal diameter, and an aneurysmal transformation of the Rex recessus (Fig. 2B). 1.2. Details of radiological assessment The Rex recessus was aneurysmal in all three cases, with the aneurysm involving the other main portal branches in two. In none was the extrahepatic portal main venous trunk found dilated. In 2 cases the latter trunk was thrombosed with secondary cavernomatous transformation. Patient 2 was remarkable with a patent portal vein and a severe stricture immediately below the bifurcation (Fig. 2A). In all cases, splenic vein, mesenteric vein, and their confluence, were normal. The maximum diameter of Rex recessus, as measured at Doppler US, ranged from 14.58 mm to 20.20 mm (mean 17.62 mm). All PVAs
437
had a fusiform shape, and at US typically presented as large anechoic structures (Fig. 1B). Colour Doppler US and duplex Doppler US showed the characteristic aspect of venous structures with constantly rotating blood flow within the lesion and its typical spectral analysis pattern in patients with high flow [8,12,14,15]. Flow in liveraneurysms was turbulent and velocities at that level ranged from 25 cm/s to 70 cm/s (mean 50.3 cm/s). On CT, the aneurysmal segments appeared as a well-circumscribed enhanced structure as part of the intrahepatic portal venous system (Fig. 2B). On T1 and T2 weighted MRI images, PVA appeared hypointense. More detailed anatomical assessment of the intrahepatic portal system was obtained in these patients by 3-dimensional CT reconstructions (Fig. 2A), and direct or indirect angiographies: this clearly delineated the aneurysmal transformation and its relation with the Rex recessus and the normal segmental portal vein branches. 1.3. Management All three patients were successfully managed surgically by the same senior surgeon: a meso-Rex bypass was performed in two, using the left internal jugular vein as a bypass (the technique has been previously described [16]). In the third case, the portal vein was freed and divided below the bifurcation: at that point, it appeared technically easier to mobilize it anteriorly and anastomose it directly, end-to-side, into the dilated Rex recessus. In all 3 patients, a high blood flow – from 40 cm/s to 50 cm/s (mean 48 cm/s) – was recorded intra-operatively within the bypass (or the portal vein in case 2), with excellent flow restored within the whole liver (Fig. 3). Current follow-up ranges from 14 to 24 months (median+SD: 18 ± 5 months; mean 18.67 months). Clinical and radiological assessments have confirmed the good outcome in all 3 patients with resolution of portal hypertension symptoms. At follow up Doppler US, the bypass flow ranged from 30 to 50 cm/s (mean 43.33 cm/s) and the aspect of aneurysm significantly normalized with a mean diameter of 15.30 mm (ranging from 13.52 mm to 18.85 mm) and a mean flow of 29.33 cm/s (ranging from 25 cm/s to 33 cm/s). The platelet count increased with time (Table 1) and the spleen is clinically no longer appreciable (spleen span at US decreased to a mean of 13.5 cm). Coagulation tests tended to normalize with time in all three patients (Table 1). No endoscopic control has been done after the operation, as we do usually after successful Meso-Rex bypass
Fig. 1. (A) Aneurysm of the Recessus of Rex (Patient II): aspect at conventional angiography (venous phase of superior mesenteric artery angiography). (B) Ultrasound aspect of an aneurysm of the right and left portal branches, typically presenting as a large anechoic fusiform structure (Patient II).
438
A. Bertocchini et al. / Journal of Pediatric Surgery 49 (2014) 436–440
Fig. 2. (A) 3D CT reconstruction of an aneurysm of the right and left portal branches, with patent portal vein trunk and a severe stricture at the bifurcation (patient II). (B) CT scan imaging of PVA: a well-circumscribed enhanced structure (Patient III). (C) 3D CT reconstruction after meso-Rex bypass. A thinner persistent stenotic native portal vein trunk is already evident (patient II). (D) Ultrasound Doppler aspect of the MesoRex bypass and PVA after operation: high velocity within the shunt and revascularization of the intrahepatic venous system (Patient II).
when the clinical condition improves and if excellent blood flow is measured in the bypass. 2. Discussion The portal vein is the most common site of visceral venous aneurysms and can present as a focal saccular or fusiform dilatation [8,17,18]. Precise etiology is still a matter of debate with investigators favoring two hypotheses. Embryogenesis and formation of the normal portal system are the results of a complex process initiated by formation venous systems (umbilical and omphalomesenteric – or vitelline – veins) bilaterally (right and left), followed by their interconnections (R–L anastomoses) and ending in the creation of the usual portal system by a final process of involution/regression of parts of the former venous system (http:// www.embryology.ch/anglais/pcardio/venen02.html and http://fr. slideshare.net/ananthatiger/embryology-vascular-development). The “congenital” theory postulates that incomplete regression of some vein or a variant branching pattern of the portal vein may determine aneurysmal transformation of the portal vein [3,19] or of the umbilical portion of the left portal vein [11]. For the supporters of the “acquired” theory, PVAs are the consequence of a weakening of the venous walls due to acute pancreatitis, or trauma, or even a consequence of a portal hypertension secondary to liver disease [3,4,19]. Not to mention that the “weakened venous wall hypothesis” has also been considered by the
supporters of the congenital theory as a possible congenital alteration of the venous wall which could cause subsequent venous dilatation under normal portal pressure [15]. Although portal hypertension is generally accepted as a cause of acquired portal vein aneurysm [12,14], the latter is rather a moderate dilatation of a patent portal vein trunk, then a true aneurysm being the consequence of a high splanchnic flow associated with hyperdynamic circulation in the cirrhotic patient. A similar effect on portal vein diameter is observed in patients with normal liver and a high portal flow secondary to arterio-portal fistula or porto-hepatic shunting [20,21]. Remarkably, in our three cases, no loco-regional abnormal condition was found and the liver was normal: there is no reason to believe that portal flow was increased as a primary cause. Overall, the various theories proposed for the genesis of PVA offer a reasonable explanation for cases with dilatation of the main portal vein, with or without venous abnormal anatomy, but not a clear explanation for those aneurysms that are located higher and especially the intrahepatic types. Also, it does not clearly explain the association between aneurysm and upstream portal cavernoma in young children. In two of our patients, the portal vein trunk was thrombosed with the aneurysm located downstream at the level of the right and left portal branches. In fact, the relation between PVA and venous thrombosis, not uncommonly seen in cirrhotic patients (up to 30 % [14]), is a “chicken–egg” debate. Turbulence or areas of blood stasis within an aneurysm could trigger thrombus formation. In these
A. Bertocchini et al. / Journal of Pediatric Surgery 49 (2014) 436–440
439
Fig. 3. (A) Aneurysm of the Recessus of Rex: aspect at indirect portography (Patient I). (B) Ultrasound aspect of a severe stricture of the main portal trunk at the level of the bifurcation (patient II). (C) Ultrasound aspect of the aneurysm of both right and left portal branches, including a huge aneurysmal transformation of the Rex recessus, typically presenting as a large anechoic fusiform structure (patient I). (D) Ultrasound Doppler aspect of the MesoRex bypass and PVA after operation: high velocity within the shunt and revascularization of the intrahepatic venous system (patient I).
According to Kimura et al., a thrombotic origin of the stricture cannot be ruled out. Interestingly, the association of intrahepatic PVA and extrahepatic portal hypertension caused by thrombosis of the portal vein trunk, in the presence of a normal liver and in the absence of extensive splanchnic venous thrombosis, makes that particular condition amenable to a cure by surgically restoring the portal system continuity. In these patients, it was achieved by direct transposition and anastomosis of the portal trunk into the dilated Rex recessus in one case, and by creating a bypass between the mesenteric vein and the Rex recessus (bypass Meso-Rex) in the two patients with portal cavernoma. The technical aspect of that procedure, as well as the fact that it respects the physiology of portal venous return through the liver, has been emphasized previously [16]. In the 3 patients, the procedure has been successful and associated with the relief of symptoms and complications caused by portal hypertension.
patients, thrombosis is a consequence, rather than the cause of the malformation, and is then located at the site of the aneurysm, not upstream as in our cases. Interestingly Kimura et al. have described in patients with thrombophilia, the formation of thrombus that may grow into an incomplete membrane: this process could lead not only to thrombosis but also to the formation of a membrane within the vein [22,23]. Patients 2 and 3 had very similar transformation of the portal venous system, the latter patient having a cavernomatous transformation of the portal vein, and the former a patent portal vein trunk with a severe stricture at the vein bifurcation. The latter stricture is evidently the cause of a typical post-stenotic dilatation of the vein. The PVA anatomy of both cases is so similar that it suggests that patient 3 had initially a similar portal vein stricture to that of Patient 2, complicated later by secondary thrombosis of the portal trunk and cavernoma formation at its level.
Table 1 Changes in spleen span and hypersplenism: before (pre) and after (post) meso-Rex bypass. FU months Patient 1 Patient 2 Patient 3
18 24 14
Spleen span (⁎)
PLT
PTT
PT
INR
Pre
Post
Pre
Post
Pre
Post
Pre
Post
Pre
Post
20 (3) 14.5 (4) 16 (5)
14.5 (0) 12 (0) 14 (0)
57 86 56
105 135 142
31.5 35.8 36.5
27.6 31.2 31.9
16 17.8 15.8
13.9 15.8 12.6
1.26 1.51 1.25
1.01 1.32 1.00
Abbreviations: FU: Follow-up, PLT: platelet count (million/ml); PTT: Partial thromboplastin time (seconds); PT: Prothrombin time (seconds); INR: International Normalized Ratio. ⁎ Spleen span was measured in cm, as maximum span at ultrasound imaging (below rib margin at clinical examination.
440
A. Bertocchini et al. / Journal of Pediatric Surgery 49 (2014) 436–440
3. Conclusions These three observations strongly support the hypothesis that a stricture, either caused by congenital stricture or partial thrombosis of the main portal vein is the primary lesion, with aneurysmal transformation being secondary to the stricture in the portal vein (as observed in case 3 and suspected in case 2). This produces turbulences and aneurysmal transformation. As a third event, complete thrombosis of the portal vein trunk may occur with cavernomatous transformation being the last event; this sequence would then explain easily that, in some patients, PVA can be found with complete thrombosis, the latter occurring at a later stage as a last complication and causing in turn a cavernomatous transformation of the portal main trunk. It is interesting to note that aneurysmal transformation of the portal vein, per se and in the absence of secondary complications (i.e. thrombosis) is not clinically relevant and, per se, does not need treatment. In contrast, prehepatic portal hypertension, presented here, as a secondary is the cause of delayed morbidity and can be cured by appropriate intervention. This series shows that the MesoRex bypass can offer a cure for those patients when the liver is normal. References [1] Ohnishi K, Nakayama T, Saito M, et al. Aneurysm of intrahepatic branch of the portal vein. Report of two cases. Gastroenterology 1984;86:169–73. [2] Koc Z, Oguzkurt L, Ulusan S. Portal venous system aneurysms: imaging, clinical findings and a possible new etiologic factor. AJR Am J Roentgenol 2007;189:1023–30. [3] Sfyroeras GS, Antoniou GA, Drakou AA, et al. Visceral venous aneurysms: clinical presentation, natural history and their management: a systematic review. Eur J Vasc Endovasc Surg 2009;38:498–505. [4] Cho SW, Marsch JW, Fontes PA, et al. Extrahepatic portal vein aneurysm — report of six patients and review of the literature. J Gastrointest Surg 2008;12:145–52. [5] Turner KC, Bohannon WT, Atkins MD. Portal vein aneurysm: a rare occurrence. J Vasc Nurs 2011;29:135–8.
[6] Yukawam N, Takahashi M, Sasaki K et al. Ultrasonography and 3D-CT follow-up of extrahepatic portal vein aneurysm: a case report. Case Report med 2010 [Epub ahead of print]. [7] Estroff JA, Benacerraf Br. Fetal umbilical vein varix: sonographic appearance and postnatal outcome. J Ultrasound Med 1992;11:69–72. [8] Haddad A, Fraiman M, Mackey R, et al. Portal vein aneurysm. Am Surgeon 2011;77(4):503–5. [9] Ohnami Y, Ishida H, Konno K, et al. Portal vein aneurysm: report of six cases and review of the literature. Abd Imaging 1997;22:281–6. [10] Flis V, Gadzijev E. Reconstruction of the main portal vein for a large saccular aneurysm. HPB (Oxford) 2003;5(3):180–90. [11] Ozbek SS, Killi MR, Pourbagher MA, et al. Portal venous system aneurysms: report of five cases. J Ultrasound Med 1999;18:417–22. [12] Yang DM, Yoon MH, Kim HS, et al. Portal vein aneurysm of the umbilical portion: imaging features and the relationship with portal vein anomalies. Abdom Imaging 2003;28:62–7. [13] Lopez-Machado E, Mallorquin-Jimenez F, Medina-Benitez A, et al. Aneurysms of the portal venous system: ultrasonography and CT findings. Eur J Radiol 1998;26: 210–4. [14] Gallego C, Velasco M, Marcuello P, et al. Congenital and acquired anomalies of the portal venous system. RadioGraphics 2002;22:141–59. [15] Atasoy KC, Fitoz S, Akyar G, et al. Aneurysm of the portal venous system gray-scale and color Doppler. Clin Imag 1998;22:414–7. [16] Alberti D, Clapuyt P, de Ville de Goyet J, et al. Direct bypassing of extrahepatic portal venous obstruction in children: a new technique for combined hepatic portal revascularization and treatment of extrahepatic portal hypertension. J Pediatr Surg 1998;33(4):597–601. [17] Dognini L, Carreri AL, Moscatelli G. Aneurysm of the portal vein: ultrasound and computed tomography identification. J Clin Ultrasound 1991;19:178–82. [18] Brock PA, Jordan Jr PH, Barth MH, et al. Portal vein aneurysm: a rare but important vascular condition. Surgery 1997;121:105–8. [19] Feliciano PD, Cullen JJ, Corson JD, et al. The management of extrahepatic portal vein aneurysms: observe or treat? HPB Surg 1996;10:113–6. [20] Laumonier H, Montaudon M, Corneloup O, et al. CT angiography of intrahepatic portal aneurysm. Abdom Imaging 2005;30:755–7. [21] Fulcher A, Turner M. Aneurysms of the portal vein and superior mesenteric vein. Abdom Imaging 1997;22:287–92. [22] Kimura K, Okuda K, Nomura F, et al. Transformation of the portal vein. Gastroenterology 1985;88:571–5. [23] Bayractar Y, Oksuzoglu G, Balkanci F, et al. Portal hypertension due to incomplete membranous obstruction of the portal vein. J Clin Gastroenterol 1995;21(3): 260–2.
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Prehepatic portal hypertension with aneurysm of the portal vein: Unusual but treatable malformative pattern Arianna Bertocchini a,⁎, Giuseppe d’Ambrosio a, Chiara Grimaldi a, Laura del Prete a, Fabrizio di Francesco a, Piergiorgio Falappa b, Lidia Monti c, Jean de Ville de Goyet a a b c
Hepato-biliary and Transplant Surgery Unit, Department of Surgery and Transplantation Centre, Bambino Gesù Childrens Hospital, Roma, Italy Interventional Radiology Unit, Department of Surgery and Transplantation Centre, Bambino Gesù Childrens Hospital, Roma, Italy Hepato-biliary Radiology Unit, Department of Imaging, Bambino Gesù Childrens Hospital, Roma, Italy
a r t i c l e
i n f o
Article history: Received 30 March 2013 Received in revised form 9 June 2013 Accepted 10 June 2013 Key words: Intrahepatic portal vein Vein Aneurysm Rex recessus Extrahepatic portal hypertension Meso Rex bypass
a b s t r a c t Introduction: Portal vein aneurysms (PVAs) are usually located at the vein trunk or at its bifurcation, rarely intra-hepatic, or at the umbilical portion. Etiology remains unclear. Methods: Three children with PVA were identified over a 5-year period. PVA anatomy was assessed by Doppler Ultrasound, Angio CT/MRI, and trans-jugular retrograde portography. Results: Three children with intrahepatic PVA (including the umbilical portion) were identified during assessment for pre-hepatic portal hypertension: all had splenomegaly and hypersplenism. One presented with massive variceal bleeding. In two cases, a portal vein cavernoma was found, and in the third a severe stricture at the portal bifurcation was observed. Restoration of portal venous flow was achieved by a mesoRex bypass in two cases and transposing the PV into the Rex in one. High hepatopetal portal flow was restored immediately, with follow-up confirming long-term patency and resolution of signs of portal hypertension with time. Conclusions: These original observations suggest a common initial malformative pattern consisting of a portal venous stricture/web causing a post-stenotic aneurysmal dilatation of the intrahepatic portal branches complicated by thrombosis and cavernomatous transformation of the portal vein trunk. Importantly, the Meso-Rex bypass allows restoring a normal portal flow and cures the portal hypertension. © 2014 Elsevier Inc. All rights reserved.
Portal vein aneurysm (PVA) is a rare vascular malformation (0.067% incidence) [1], representing fewer than 3% of all venous aneurysms [2]. First described by Barziali and Kleckner in 1956 [3,4], almost 150 cases are now reported in literature [5] (only 68 cases in English) [6]). Morphologically, portal vein aneurysm can be fusiform or saccular [7]. Two main categories have been described: intrahepatic or extrahepatic. Extrahepatic aneurysms are more common, located at the main portal vein in 52% of cases, at the superior mesenteric– splenic vein confluence in 44%, and rarely (3% to 4%) at the level of the right or left portal branches [4]. Portal vein diameter being variable, an aneurysmal dilatation of the PV is defined according to the venous upper limit of range for age [8–10]. Intrahepatic aneurysms and those involving the umbilical portion of portal vein are rare with only few cases being reported [11–13]. Clinical presentation varies, especially in relation to the aneurysm size, the majority of patients presenting with nonspecific abdominal pain (44% to 54%), while many others are asymptomatic and diagnosis ⁎ Corresponding author. Department of Pediatric Surgery and Transplantation Center, Bambino Gesù Childrens Hospital, Piazza Sant’Onofrio 4, 00165 Roma, Italy. Tel.: +39 0668592851; +390668592425; +393383203451; fax: +390668593841. E-mail address: [email protected] (A. Bertocchini). 0022-3468/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpedsurg.2013.06.012
is incidental (25% to 38%). Rarely, patients are diagnosed with portal hypertension or gastrointestinal bleeding (7.3% to 9.8%) [3,4]. In this report, 3 children with pre-hepatic portal hypertension and intrahepatic portal aneurysms are presented, and a surgical cure is described. 1. Patients and methods Between January 2007 and December 2011, three children with PVA were identified out of a population of 105 children who were radiologically evaluated for extrahepatic portal hypertension (3%, all 3 female). Patient characteristics, clinical presentation, anatomy of the PVA, surgical correction and outcome were reviewed. Anatomical assessment of the anatomy was done by standard imaging (Doppler Ultrasound (US), Angio CT, Angio MRI) and transjugular hepatic vein catheterization (gradient measurement and retrograde portography) in all cases. 1.1. Patients 1.1.1. Patient 1 At the age of 18, she was admitted to our centre after a single episode of massive variceal bleeding a month earlier. Endoscopy had shown four F2 varices with red stripes in the lower portion of the
A. Bertocchini et al. / Journal of Pediatric Surgery 49 (2014) 436–440
esophagus, and she was treated by elastic bands positioning. She had splenomegaly (US span: 20 cm), thrombocytopenia (platelet count: 57,000/ml) and normal hepatic function tests. At radiological assessment, a portal cavernoma and aneurysm of both right and left portal branches, including a huge aneurysmal transformation of the Rex recessus, were observed. 1.1.2. Patient 2 A 5-year-old girl presented initially with variceal bleeding; esophageal varices, factor VII deficiency, splenomegaly (US span: 14.5 cm) and thrombocytopenia (platelet count: 86,000/ml). At imaging, a malformative pattern combining PVA of both portal branches extending into the Rex recessus, a normal extrahepatic portal vein trunk and a severe stricture of the main portal trunk at the level of the bifurcation were found (Figs. 1 and 2A). An attempt at percutaneous transhepatic angioplasty/stenting of the stricture in another centre failed, and she was referred to our centre for surgical treatment. 1.1.3. Patient 3 Splenomegaly and portal vein cavernoma were incidentally diagnosed in a 3-year old girl with an uneventful history and absence of previous gastrointestinal bleeding. Endoscopy showed esophageal varices F2, gastric varices and mild hypertensive gastropathy. She was in good general conditions, with splenomegaly (US span: 16 cm), thrombocytopenia (platelet count: 56,000/ml) and normal hepatic function tests. Radiological assessment showed a portal cavernoma with patent intrahepatic portal branches of a normal diameter, and an aneurysmal transformation of the Rex recessus (Fig. 2B). 1.2. Details of radiological assessment The Rex recessus was aneurysmal in all three cases, with the aneurysm involving the other main portal branches in two. In none was the extrahepatic portal main venous trunk found dilated. In 2 cases the latter trunk was thrombosed with secondary cavernomatous transformation. Patient 2 was remarkable with a patent portal vein and a severe stricture immediately below the bifurcation (Fig. 2A). In all cases, splenic vein, mesenteric vein, and their confluence, were normal. The maximum diameter of Rex recessus, as measured at Doppler US, ranged from 14.58 mm to 20.20 mm (mean 17.62 mm). All PVAs
437
had a fusiform shape, and at US typically presented as large anechoic structures (Fig. 1B). Colour Doppler US and duplex Doppler US showed the characteristic aspect of venous structures with constantly rotating blood flow within the lesion and its typical spectral analysis pattern in patients with high flow [8,12,14,15]. Flow in liveraneurysms was turbulent and velocities at that level ranged from 25 cm/s to 70 cm/s (mean 50.3 cm/s). On CT, the aneurysmal segments appeared as a well-circumscribed enhanced structure as part of the intrahepatic portal venous system (Fig. 2B). On T1 and T2 weighted MRI images, PVA appeared hypointense. More detailed anatomical assessment of the intrahepatic portal system was obtained in these patients by 3-dimensional CT reconstructions (Fig. 2A), and direct or indirect angiographies: this clearly delineated the aneurysmal transformation and its relation with the Rex recessus and the normal segmental portal vein branches. 1.3. Management All three patients were successfully managed surgically by the same senior surgeon: a meso-Rex bypass was performed in two, using the left internal jugular vein as a bypass (the technique has been previously described [16]). In the third case, the portal vein was freed and divided below the bifurcation: at that point, it appeared technically easier to mobilize it anteriorly and anastomose it directly, end-to-side, into the dilated Rex recessus. In all 3 patients, a high blood flow – from 40 cm/s to 50 cm/s (mean 48 cm/s) – was recorded intra-operatively within the bypass (or the portal vein in case 2), with excellent flow restored within the whole liver (Fig. 3). Current follow-up ranges from 14 to 24 months (median+SD: 18 ± 5 months; mean 18.67 months). Clinical and radiological assessments have confirmed the good outcome in all 3 patients with resolution of portal hypertension symptoms. At follow up Doppler US, the bypass flow ranged from 30 to 50 cm/s (mean 43.33 cm/s) and the aspect of aneurysm significantly normalized with a mean diameter of 15.30 mm (ranging from 13.52 mm to 18.85 mm) and a mean flow of 29.33 cm/s (ranging from 25 cm/s to 33 cm/s). The platelet count increased with time (Table 1) and the spleen is clinically no longer appreciable (spleen span at US decreased to a mean of 13.5 cm). Coagulation tests tended to normalize with time in all three patients (Table 1). No endoscopic control has been done after the operation, as we do usually after successful Meso-Rex bypass
Fig. 1. (A) Aneurysm of the Recessus of Rex (Patient II): aspect at conventional angiography (venous phase of superior mesenteric artery angiography). (B) Ultrasound aspect of an aneurysm of the right and left portal branches, typically presenting as a large anechoic fusiform structure (Patient II).
438
A. Bertocchini et al. / Journal of Pediatric Surgery 49 (2014) 436–440
Fig. 2. (A) 3D CT reconstruction of an aneurysm of the right and left portal branches, with patent portal vein trunk and a severe stricture at the bifurcation (patient II). (B) CT scan imaging of PVA: a well-circumscribed enhanced structure (Patient III). (C) 3D CT reconstruction after meso-Rex bypass. A thinner persistent stenotic native portal vein trunk is already evident (patient II). (D) Ultrasound Doppler aspect of the MesoRex bypass and PVA after operation: high velocity within the shunt and revascularization of the intrahepatic venous system (Patient II).
when the clinical condition improves and if excellent blood flow is measured in the bypass. 2. Discussion The portal vein is the most common site of visceral venous aneurysms and can present as a focal saccular or fusiform dilatation [8,17,18]. Precise etiology is still a matter of debate with investigators favoring two hypotheses. Embryogenesis and formation of the normal portal system are the results of a complex process initiated by formation venous systems (umbilical and omphalomesenteric – or vitelline – veins) bilaterally (right and left), followed by their interconnections (R–L anastomoses) and ending in the creation of the usual portal system by a final process of involution/regression of parts of the former venous system (http:// www.embryology.ch/anglais/pcardio/venen02.html and http://fr. slideshare.net/ananthatiger/embryology-vascular-development). The “congenital” theory postulates that incomplete regression of some vein or a variant branching pattern of the portal vein may determine aneurysmal transformation of the portal vein [3,19] or of the umbilical portion of the left portal vein [11]. For the supporters of the “acquired” theory, PVAs are the consequence of a weakening of the venous walls due to acute pancreatitis, or trauma, or even a consequence of a portal hypertension secondary to liver disease [3,4,19]. Not to mention that the “weakened venous wall hypothesis” has also been considered by the
supporters of the congenital theory as a possible congenital alteration of the venous wall which could cause subsequent venous dilatation under normal portal pressure [15]. Although portal hypertension is generally accepted as a cause of acquired portal vein aneurysm [12,14], the latter is rather a moderate dilatation of a patent portal vein trunk, then a true aneurysm being the consequence of a high splanchnic flow associated with hyperdynamic circulation in the cirrhotic patient. A similar effect on portal vein diameter is observed in patients with normal liver and a high portal flow secondary to arterio-portal fistula or porto-hepatic shunting [20,21]. Remarkably, in our three cases, no loco-regional abnormal condition was found and the liver was normal: there is no reason to believe that portal flow was increased as a primary cause. Overall, the various theories proposed for the genesis of PVA offer a reasonable explanation for cases with dilatation of the main portal vein, with or without venous abnormal anatomy, but not a clear explanation for those aneurysms that are located higher and especially the intrahepatic types. Also, it does not clearly explain the association between aneurysm and upstream portal cavernoma in young children. In two of our patients, the portal vein trunk was thrombosed with the aneurysm located downstream at the level of the right and left portal branches. In fact, the relation between PVA and venous thrombosis, not uncommonly seen in cirrhotic patients (up to 30 % [14]), is a “chicken–egg” debate. Turbulence or areas of blood stasis within an aneurysm could trigger thrombus formation. In these
A. Bertocchini et al. / Journal of Pediatric Surgery 49 (2014) 436–440
439
Fig. 3. (A) Aneurysm of the Recessus of Rex: aspect at indirect portography (Patient I). (B) Ultrasound aspect of a severe stricture of the main portal trunk at the level of the bifurcation (patient II). (C) Ultrasound aspect of the aneurysm of both right and left portal branches, including a huge aneurysmal transformation of the Rex recessus, typically presenting as a large anechoic fusiform structure (patient I). (D) Ultrasound Doppler aspect of the MesoRex bypass and PVA after operation: high velocity within the shunt and revascularization of the intrahepatic venous system (patient I).
According to Kimura et al., a thrombotic origin of the stricture cannot be ruled out. Interestingly, the association of intrahepatic PVA and extrahepatic portal hypertension caused by thrombosis of the portal vein trunk, in the presence of a normal liver and in the absence of extensive splanchnic venous thrombosis, makes that particular condition amenable to a cure by surgically restoring the portal system continuity. In these patients, it was achieved by direct transposition and anastomosis of the portal trunk into the dilated Rex recessus in one case, and by creating a bypass between the mesenteric vein and the Rex recessus (bypass Meso-Rex) in the two patients with portal cavernoma. The technical aspect of that procedure, as well as the fact that it respects the physiology of portal venous return through the liver, has been emphasized previously [16]. In the 3 patients, the procedure has been successful and associated with the relief of symptoms and complications caused by portal hypertension.
patients, thrombosis is a consequence, rather than the cause of the malformation, and is then located at the site of the aneurysm, not upstream as in our cases. Interestingly Kimura et al. have described in patients with thrombophilia, the formation of thrombus that may grow into an incomplete membrane: this process could lead not only to thrombosis but also to the formation of a membrane within the vein [22,23]. Patients 2 and 3 had very similar transformation of the portal venous system, the latter patient having a cavernomatous transformation of the portal vein, and the former a patent portal vein trunk with a severe stricture at the vein bifurcation. The latter stricture is evidently the cause of a typical post-stenotic dilatation of the vein. The PVA anatomy of both cases is so similar that it suggests that patient 3 had initially a similar portal vein stricture to that of Patient 2, complicated later by secondary thrombosis of the portal trunk and cavernoma formation at its level.
Table 1 Changes in spleen span and hypersplenism: before (pre) and after (post) meso-Rex bypass. FU months Patient 1 Patient 2 Patient 3
18 24 14
Spleen span (⁎)
PLT
PTT
PT
INR
Pre
Post
Pre
Post
Pre
Post
Pre
Post
Pre
Post
20 (3) 14.5 (4) 16 (5)
14.5 (0) 12 (0) 14 (0)
57 86 56
105 135 142
31.5 35.8 36.5
27.6 31.2 31.9
16 17.8 15.8
13.9 15.8 12.6
1.26 1.51 1.25
1.01 1.32 1.00
Abbreviations: FU: Follow-up, PLT: platelet count (million/ml); PTT: Partial thromboplastin time (seconds); PT: Prothrombin time (seconds); INR: International Normalized Ratio. ⁎ Spleen span was measured in cm, as maximum span at ultrasound imaging (below rib margin at clinical examination.
440
A. Bertocchini et al. / Journal of Pediatric Surgery 49 (2014) 436–440
3. Conclusions These three observations strongly support the hypothesis that a stricture, either caused by congenital stricture or partial thrombosis of the main portal vein is the primary lesion, with aneurysmal transformation being secondary to the stricture in the portal vein (as observed in case 3 and suspected in case 2). This produces turbulences and aneurysmal transformation. As a third event, complete thrombosis of the portal vein trunk may occur with cavernomatous transformation being the last event; this sequence would then explain easily that, in some patients, PVA can be found with complete thrombosis, the latter occurring at a later stage as a last complication and causing in turn a cavernomatous transformation of the portal main trunk. It is interesting to note that aneurysmal transformation of the portal vein, per se and in the absence of secondary complications (i.e. thrombosis) is not clinically relevant and, per se, does not need treatment. In contrast, prehepatic portal hypertension, presented here, as a secondary is the cause of delayed morbidity and can be cured by appropriate intervention. This series shows that the MesoRex bypass can offer a cure for those patients when the liver is normal. References [1] Ohnishi K, Nakayama T, Saito M, et al. Aneurysm of intrahepatic branch of the portal vein. Report of two cases. Gastroenterology 1984;86:169–73. [2] Koc Z, Oguzkurt L, Ulusan S. Portal venous system aneurysms: imaging, clinical findings and a possible new etiologic factor. AJR Am J Roentgenol 2007;189:1023–30. [3] Sfyroeras GS, Antoniou GA, Drakou AA, et al. Visceral venous aneurysms: clinical presentation, natural history and their management: a systematic review. Eur J Vasc Endovasc Surg 2009;38:498–505. [4] Cho SW, Marsch JW, Fontes PA, et al. Extrahepatic portal vein aneurysm — report of six patients and review of the literature. J Gastrointest Surg 2008;12:145–52. [5] Turner KC, Bohannon WT, Atkins MD. Portal vein aneurysm: a rare occurrence. J Vasc Nurs 2011;29:135–8.
[6] Yukawam N, Takahashi M, Sasaki K et al. Ultrasonography and 3D-CT follow-up of extrahepatic portal vein aneurysm: a case report. Case Report med 2010 [Epub ahead of print]. [7] Estroff JA, Benacerraf Br. Fetal umbilical vein varix: sonographic appearance and postnatal outcome. J Ultrasound Med 1992;11:69–72. [8] Haddad A, Fraiman M, Mackey R, et al. Portal vein aneurysm. Am Surgeon 2011;77(4):503–5. [9] Ohnami Y, Ishida H, Konno K, et al. Portal vein aneurysm: report of six cases and review of the literature. Abd Imaging 1997;22:281–6. [10] Flis V, Gadzijev E. Reconstruction of the main portal vein for a large saccular aneurysm. HPB (Oxford) 2003;5(3):180–90. [11] Ozbek SS, Killi MR, Pourbagher MA, et al. Portal venous system aneurysms: report of five cases. J Ultrasound Med 1999;18:417–22. [12] Yang DM, Yoon MH, Kim HS, et al. Portal vein aneurysm of the umbilical portion: imaging features and the relationship with portal vein anomalies. Abdom Imaging 2003;28:62–7. [13] Lopez-Machado E, Mallorquin-Jimenez F, Medina-Benitez A, et al. Aneurysms of the portal venous system: ultrasonography and CT findings. Eur J Radiol 1998;26: 210–4. [14] Gallego C, Velasco M, Marcuello P, et al. Congenital and acquired anomalies of the portal venous system. RadioGraphics 2002;22:141–59. [15] Atasoy KC, Fitoz S, Akyar G, et al. Aneurysm of the portal venous system gray-scale and color Doppler. Clin Imag 1998;22:414–7. [16] Alberti D, Clapuyt P, de Ville de Goyet J, et al. Direct bypassing of extrahepatic portal venous obstruction in children: a new technique for combined hepatic portal revascularization and treatment of extrahepatic portal hypertension. J Pediatr Surg 1998;33(4):597–601. [17] Dognini L, Carreri AL, Moscatelli G. Aneurysm of the portal vein: ultrasound and computed tomography identification. J Clin Ultrasound 1991;19:178–82. [18] Brock PA, Jordan Jr PH, Barth MH, et al. Portal vein aneurysm: a rare but important vascular condition. Surgery 1997;121:105–8. [19] Feliciano PD, Cullen JJ, Corson JD, et al. The management of extrahepatic portal vein aneurysms: observe or treat? HPB Surg 1996;10:113–6. [20] Laumonier H, Montaudon M, Corneloup O, et al. CT angiography of intrahepatic portal aneurysm. Abdom Imaging 2005;30:755–7. [21] Fulcher A, Turner M. Aneurysms of the portal vein and superior mesenteric vein. Abdom Imaging 1997;22:287–92. [22] Kimura K, Okuda K, Nomura F, et al. Transformation of the portal vein. Gastroenterology 1985;88:571–5. [23] Bayractar Y, Oksuzoglu G, Balkanci F, et al. Portal hypertension due to incomplete membranous obstruction of the portal vein. J Clin Gastroenterol 1995;21(3): 260–2.
