- Email: [email protected]
Hemodynamics in extrahepatic portal vein obstruction and its changes during long-term follow-ups
Hepatology Research 24 (2002) 141– 151
www.elsevier.com/locate/ihepcom
Hemodynamics in extrahepatic portal vein obstruction and its changes during long-term follow-ups Akiko Ogawa *, Shigeru Takamori, Kuniaki Kojima, Masaki Fukasawa, Tomoe Beppu, Shunji Futagawa Second Department of Surgery, School of Medicine, Juntendo Uni6ersity, 2 -1 -1 Hongo, Bunkyo-ku, Tokyo 113 -8421, Japan Received 26 November 2001; received in revised form 4 February 2002; accepted 8 February 2002
Abstract Thirty-six patients with extrahepatic portal vein obstruction (EHO) were studied. Twenty-one patients, whose age at onset was under 20 years old, were thought to be primary EHO. Ten of 15 patients over 21 years old, had a history of abdominal diseases, and were thus considered to be secondary EHO; only 5 were primary EHO. EHO was classified into three categories, based on selective celiac-superior mesenteric arteriography, and intraoperative trans-splenic and superior mesenteric venography: the hilar obstruction of portal vein, the portal trunk obstruction, and the portal trunk–splenic vein obstruction. Both primary and secondary EHO cases were found to have extended their obstructed region over time by progressing from hepatic hilum to portal trunk obstruction followed by splenic vein obstruction. On the other hand, based on retrograde portography, EHO was classified into the absence of intrahepatic portal obstruction (portal trunk type) and the intrahepatic obstruction (portal branch type) and the portal branch type tended to predominate. Changes of intra- and extra-hepatic portal obstruction were examined by comparing various angiographies of the 14 patients available for between 1 and 26 years of long-term follow-up. While EHO tended to progress with time, no patient showed progress of a intrahepatic portal vein obstruction. From these results, EHO patients are expected to show continuous progression of EHOs during long-term follow-ups, and to have high recurrence rates of gastrointestinal varices after various treatments due to the persistent portal hypertension. However, effective hepatic flow is likely to be maintained because collateral circulation to the liver develops, and intrahepatic portal vein obstruction does not progress, resulting in a good prognosis for EHO. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Portal hypertension; Cavernomatous transformation; Intrahepatic portal vein obstruction
1. Introduction Extrahepatic portal vein obstruction (EHO) is the third greatest cause of portal hypertension * Corresponding author. Tel.: +81-3-5802-1078; fax: + 813-3818-7589
following hepatic cirrhosis (LC) and idiopathic portal hypertension (IPH). While its prognosis is better than LC and IPH, EHO is characterized by a high recurrence rate of varices after various treatments and by a wide age of onset (from children to adults). Additionally, it is said that there are a wide variety of causes of portal ob-
1386-6346/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 6 - 6 3 4 6 ( 0 2 ) 0 0 0 2 9 - 3
142
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
struction. The present study examined the portal obstructed region using various angiographies, the morphology of cavernomatous transformation and their changes during long-term follow-up in the patients with EHO. 2. Subjects and methods The subjects were 36 EHO patients (23 males and 13 females, 5– 76 years old, average 38.2 years) admitted to our department since September 1979 (Table 1). In the table, we defined the age of onset as the age at which the clinical symptoms developed. EHO was diagnosed by selective celiac and superior mesenteric arteriography at the initial operation for esophagogastric varices that were preoperatively performed in all cases, and by intraoperative trans-splenic and superior mesenteric venography. We defined the site of obstruction in EHO as the defect in portography including intraoperative angiography in the present study, because it was so difficult to detect the site of obstruction accurately with intraoperative findings. As to intraoperative angiography, although trans-superior mesenteric venography was applied before 1985, trans-splenic venography has been performed in all cases excluding the cases with a history of splenectomy since we experienced a complication of portal thrombus after trans-superior mesenteric venography. From the findings of angiographs, EHO was classified into three categories (Fig. 1): (1) the hilar obstruction of portal vein type; the portal trunk opened and only the hilar region of portal vein was obstructed, (2) the portal trunk obstruction type; the entire portal trunk was obstructed and replaced with cavernomatous transformation, and (3) the portal trunk –splenic vein obstruction type; the portal trunk and splenic vein were obstructed. The findings of venographies did not be affected by the methods such as preoperative and postoperative venographies. Angiography also showed that cavernomatous transformation has been developed in all type. The cavernomatous transformations flowed almost evenly into both branches of portal vein and there was no case in which the dominant flow was found into either branch of portal vein.
The morphology of EHO was further classified into two categories by retrograde portography performed by wedging a catheter into the peripheral hepatic vein during hepatic vein catheterization (Fig. 2): (1) the absence of intrahepatic portal obstruction (portal trunk type) (extrahepatic type); intrahepatic portal vein obstruction was limited in the hepatic portal region and communication between right and left hepatic portal veins was observed, and (2) the intrahepatic obstruction (portal branch type) (intrahepatic type); obstruction extended over the intrahepatic portal vein and no communication between the right and left hepatic portal vein was observed. These categories were associated with age of onset, history of abdominal diseases, and duration of symptoms in the 36 subjects. The retrograde portography was carried out with manual injection to prevent a complication of sub-capsule hematoma of the liver. EHOs of No. 27, 28, 35 those complicated with LC or IPH were excluded from the study on retrograde portography because the contrast on the angiography was restricted by presinusoidal or postsinusoidal blocks. On the other hand, because contrast medium was injected into portal vein without much resistance, injected pressure did not affect the examination of obstructed portion in the other EHOs. The retrograde portographies were performed on 32 cases with informed consent excluding emergent cases, cases without agreement to the angiography, and cases with previous operation at other hospital. Various angiographies in during the follow-up periods were also compared with each other and the morphological changes of intra-/extra-hepatic portal vein obstruction and cavernomatous transformation over time were observed in the 14 patients who were available for 1– 26 years of follow-up study after their initial operations. In the follow-up study, we defined the extension of the defect in venography as the progression of EHO.
3. Results The age of onset of the 36 EHO patients ranged from 1 to 64 years with the most common initial symptoms of hematemesis/melena, followed by
Age of onset
1 2 4 4
4 4
4 5 7 7 10 12 14 14 15 15 16
16
16 17 19
25
27 27 32 39
48
Age and sex
11 ß 5à 16 ß 16 ß
23 ß 28 ß
40 ß 40 à 7ß 37 ß 22 à 13 à 14 ß 41 ß 17 à 37 ß 17 ß
21 à
53 ß 17 ß 53 à
33 à
ß ß ß ß
27 28 34 41
50 ß
Case
1 2 3 4
5 6
7 8 9 10 11 12 13 14 15 16 17
18
19 20 21
22
23 24 25 26
27
Table 1 Clinical features of patients with EHO
Esophagogastric varices Melena Abdominal pain Splenomegaly Esophagogastric varices Esophagogastric varices
Abdominal distension Unknown Hematemesis Nasal hemorrhage
Splenomegaly Melena Hematemesis Hematemesis and melena Splenomegaly Hematemesis and melena Splenomegaly Anemia Hematemesis Unknown Hematemesis Pancytopenia Splenomegaly Hematemesis Abdominal pain Splenomegaly Splenomegaly
Clinical findings
Ope for appendicitis (15), for peritonitis (15), for ileus (20)
Portal thrombosis (27) Ope for appendicitis(9) Ope for appendicitis (16)
Exchange transfusion via umbilical vein
Past history of abdominal diseases (age)
2
0 1 2 2
8
37 0 34
5
36 35 0 30 12 1 0 27 2 22 1
19 24
10 3 12 12
Period between onset and angiography (years)
H
S P S S
P
P H H
S
P P H P H H P H P P P
P H
H H H H
Portal phase of arteriography
I
I I I I
I
I E E
I
I I I I I I I – I I I
I I
E I E E
Retrograde portography
Site of obstruction
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151 143
49
52
54
55
55 56
61 62
64
50 à
52 à
59 ß
55 à
58 ß 57 à
65 ß 62 ß
76 à
28
29
30
31
32 33
34 35
36
Esophagogastric varices
Esophagogastric varices Point out at operation Hematemesis By ultrasonography Unknown Esophagogastric varices
Esophagogastric varices Esophagogastric varices
Clinical findings
Chronic pancreatitis (51) Ope for gastric ulcer (46), for choledocholithiasis and liver abscess (60)
Choledocholithiasis (16), cholangitis (55) Cholelithiasis (51)
Ope for cholelithiasis and choledocholithiasis (36), for peritonitis (47)
Cholecystitis (49)
Past history of abdominal diseases (age)
12
4 0
3 1
0
5
0
1
Period between onset and angiography (years)
Ope, operation; H, hilum; P, portal trunk; I, intrahepatic; S, P and splenic vein; E, extrahepatic.
Age of onset
Age and sex
Case
Table 1 Clinical features of patients with EHO
P
P H
P P
H
P
P
P
Portal phase of arteriography
–
I I
– I
–
E
I
I
Retrograde portography
Site of obstruction
144 A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
esophageal/gastric varices and splenomegaly (Table 1). Patients with any past history seemed to be a causal factor of EHO were a case who had undergone neonatal exchange transfusion via the umbilical vein, 5 cases who had cholelithiasis/ choledocholithiasis, and 5 cases who underwent abdominal operation.
3.1. Age of EHO onset and types of intra-/extra-hepatic portal obstruction Table 2 provides a summary of the site of obstruction and the age of onset. More than half
145
(11 of 21, 52.3%) patients whose age of EHO onset was under 20 years had hilar obstruction and only one patient (4.8%) had a portal trunk– splenic vein obstruction. In patients whose age of EHO onset was over 21 years, no more than 4 of 15 patients (26.7%) had hilar obstruction and 3 patients (20.0%) had portal trunk–splenic vein obstruction, showing that the EHO tends to be diffuse. In all cases, cavernomatous transformations were found to extend beyond the portal obstructive region to the liver. From the findings of retrograde portography in 31 patients with EHO the portal trunk type extrahepatic type con-
Fig. 1. Portographies show (a) hilar obstruction of portal vein type, (b) portal trunk obstruction type and (c) portal trunk – splenic vein obstruction type, in which the portal trunk is obstructed in portal phase of superior mesenteric arteriography (c-1) and splenic vein of celiac arteriography (c-2). Arrows show the part of obstruction.
146
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
Fig. 2. Retrograde portgraphies show (a) extrahepatic type (cannulated through left hepatic vein) and (b) intrahepatic type (cannulated through right hepatic vein). Arrows show the site of obstruction.
sist of 5 patients (25.0%) whose age of onset was under 20 years and a patient (11.1%) over 21 years, and the portal branch type intrahepatic type of 15 (75.0%) and 8 (88.9%), showing that the portal branch type tended to increase in the patients whose age at onset was over 21 years old (Table 3).
3.2. History of abdominal diseases and types of intra-/extra-hepatic portal obstruction In the patients whose age at EHO onset was over 21 years, 3 of 5 patients without a history of abdominal diseases had portal trunk obstruction. In patients with a history of abdominal diseases, 3 patients had hilar obstructions (30.0%), 5 had portal trunk obstructions (50.0%), and 2 had portal trunk –splenic vein obstruction (20.0%); showing a slightly lower ratio of hilar obstruction (Table 4). In 12 patients whose age at EHO onset was over 21 years and who underwent retrograde portography, of those with a history of abdominal disease, none had the portal trunk type extrahepatic type and 5 (100%) had the portal branch type intrahepatic type. In patients without a history of abdominal disease, one had portal trunk type (25.0%) extrahepatic type and 3 the portal branch type
intrahepatic type (75.0%). In both cases the numbers of patients with portal branch type intrahepatic type were higher than with portal trunk type extrahepatic type (Table 5).
Table 2 Age of onset with EHO and site of obstruction Site of obstruction
Hepatic hilum Portal trunk Portal trunk and splenic vein
Cases
15 17 4
Age of onset (years) 0–20
Over 21
11 9 1
4 8 3
Table 3 Age of onset with EHO and site of obstruction from the findings of retrograde portography Site of obstruction
Cases
Extrahepatic portal vein 6 Intrahepatic portal vein 23
Age of onset (years) 0–20
Over 21
5 15
1 8
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
147
Table 4 Past history of abdominal diseases with onset age of over 21 years and site of obstruction Site of obstruction
Cases
Abdominal diseases (−)
Abdominal diseases (+)
Hepatic hilum Portal trunk Portal trunk and splenic vein
4 8 3
1 3 1
3 5 2
Table 5 Past history of abdominal diseases with onset age of over 21 years and site of obstruction from the findings of retrograde portography Site of obstruction
Cases
Abdominal diseases (−)
Abdominal diseases (+)
Extrahepatic portal vein Intrahepatic portal vein
1 8
1 3
0 5
Table 6 Period after abdominal disease or operation at onset age with EHO of over 21 years and site of obstruction Site of obstruction
Hepatic hilum Portal trunk Portal trunk and splenic vein
Cases
3 4 2
In the patients with a history of abdominal diseases, 2 of 4 whose period from suffering the disease to angiography was within 2 years had hilar obstruction, 3 of 3 whose period was 3– 10 years had portal trunk obstruction, and 2 of 3 whose period was over 10 years had portal trunk –splenic vein obstruction. The number of patients with portal trunk and portal trunk– splenic vein obstructions tended to increase with time (Table 6).
3.3. Changes of portal obstruction in the patients subjected to long-term follow-up Various angiographies at the initial operation and during the follow-up period were compared with each other to examine the changes of intra-/ extra-hepatic portal obstruction in 14 patients
Period after abdominal disease or operation (years) Below 2
3–10
Over 10
2 1 0
0 3 0
1 0 2
who were available for long-term follow-up ranging from 1 to 26 years. Four cases of hilar obstruction showed an extension to the portal trunk obstruction, and one portal trunk obstruction case changed to the portal trunk–splenic vein obstruction. We present a case, with melena as an initial symptom, that intra-operative venography at the operation of esophageal transection for esophagogastric varices at the age of 5 and the portographies after 12 years and 16 years after operation were available in Fig. 3, and retrograde portographies at the operation and 12 years after operation were also available in Fig. 4. Intra-operative venography showed that the portal trunk opened and only the hilar region of portal vein was obstructed (hilar type). At the 12 years after operation, the portal trunk still opened. At the 16 years after operation the obstruction developed to the portal trunk, and the morphology of caver-
148
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
nomatous transformation also changed. The retrograde portography showed the obstruction was portal branch type intrahepatic type at the operation, and 12 years after operation the obstruction did not developed (Table 7). On the other hand, in the 5 patients who twice underwent retrograde portography, none showed extensions of intrahepatic portal obstructions.
4. Discussion According to the guidelines for the diagnosis of EHO made by Ministry of Health, Labour and Welfare Specified Diseases Study Group for Abnormal Portal Circulation [1], EHO can be divided into two categories: primary EHO, whose causes are unknown, and secondary EHO, whose
Fig. 3. Portographies show (a) hilar obstruction at the operation for esophagogastric varices in the intra-operative venography, (b) 12 years after operation and (c) portal trunk obstruction in the portal phase of superior mesenteric arteriography at 16 years after operation.
Fig. 4. Retrograde portographies show (a) intrahepatic obstruction at the time of operation and (b) 12 years after operation. No progression of the obstruction is found between at the time of operation and 12 years after operation.
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
149
Table 7 Changes in long-term follow-up portography Case
2 3 20 36 7 8 10 13 16 22 34 23 25 26
Age of onset
2 4 17 64 4 5 7 14 15 25 61 27 32 39
Site of obstruction (period; years) Portal phase of arteriography
Retrograde portography
H(12) H(16) P H P (3) H(10) P(12) P P H (1) P P (1) P(15) P P P (7) P(26) P P P (2) P(11) P P S (1) S S (1) S S (3) S
I(12) I H(10) H I(15) I I(22) I
I I (9)
H, hilum; P, portal trunk; I, intrahepatic; S, P and splenic vein; E, extrahepatic.
causes includes hepatic cirrhosis, IPH, tumor, hematologic disorders, cholecystitis/cholangitis, pancreatitis and abdominal operation. Some reports in recent years have noted that inherited or acquired hypercoagulable states such as protein C, protein S, and antithrombin III deficiency may predispose patients to develop portal venous obstruction [2]. As for the age distribution of EHO, Webb and Sherlock [3] reported that there were two peaks in ages of EHO onset in childhood and adulthood. The Specified Diseases Study Group for Abnormal Portal Circulation also reported from national statistics [4] that a clear peak of EHO onset was found in teenagers and another slight peak was found in the forties in Japan. Since most child patients did not have any causal past history resulting in secondary EHO, several authors have suggested an association between EHO and certain congenital anomalies such as ventricular septal defect, atrial septal defect, deformed inferior vena cava and biliary tract anomalies [5] [6]. Recently, Ando et al. [7] suggested the abnormal running of portal vein, an anomaly during development, as a potential cause. A subsequent study reported that the most common etiologic factor in children is infection, accounting for 43–52% of all cases [8], and umbil-
ical sepsis, the most frequent infection during the neonatal period, is present in 10– 26% of children [9]. One patient in this study whose age at EHO onset was under 20 years had a history of exchange transfusion via the umbilical vein. While there was no history of abdominal diseases in patients whose age of onset was under 20 years, 10 of 15 patients whose age at onset was over 21 years (66.7%) had histories of abdominal diseases. Five of these underwent abdominal operations mainly for the treatment of appendicitis and cholecystolithiasis/choledocholithiasis. Most of appendicitis in these patients had progressed to peritonitis. Five cases were likely to progress from LC or IPH. Five of the 15 patients whose age at onset was over 21 years were thought to have primary EHO, but most were suggested to have so-called secondary EHO. The present study examined the morphology of cavernomatous transformation complicated with EHO. Hilar obstruction was slightly predominant in patients whose age at EHO onset was under 20 years, while extensive obstructions such as portal trunk obstruction including the splenic vein was predominant in patients whose age at EHO onset was over 21 years old. As for primary EHO, hilar obstruction was predominant in patients whose
150
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
age at onset was under 20 years. No hilar obstruction was observed in the patients whose age at onset was over 21 years. Since the ratio of portal trunk obstruction became higher as the age at onset progressed, portal obstruction is suggested to progress from hepatic hilum to portal trunk. Obstruction was found in various regions in secondary EHO. In an examination of the regions of the obstruction based on the period since abdominal disease or operation, EHO tended to progress from hilar obstruction to portal trunk obstruction, and further to splenic vein obstruction over time. Thus both primary and secondary EHO tend to extend their obstruction ranges as years of suffering the condition increase. It has been reported that the most common initial symptom of childhood EHO was hematemesis/melena followed by splenomegaly [3] and the same tendency has been observed in our department. Only one patient had hematemesis as a initial symptom in the patients whose age at primary EHO onset was over 21 years. Cavernomatous transformation to the liver has long been reported. Beginning with Balfour and Stewart in 1868 [10] followed by Klemperer in 1928 who reported additional similar cases [11] and named it cavernomatous transformation based on the histopathological morphology of the hepatic portal region. Nowadays, causal factors of cavernomatous transformation are considered as the recanalization of the portal vein due to portal trunk obstruction and dilated collateral circulation to the liver. Previously it was thought that long-trunk portal obstruction caused cavernomatous transformation [12– 14]. However, Honna [15] reported that the gradual obstruction of the portal vein induced by an ameroid constrictor resulted in development of collateral circulation to/from the liver along with portal hypertension in rats 9 weeks after constrictor application. Ibuki et al. [16] also reported that cavernomatous transformation was completely established 3 months after the development of portal thrombosis. Moreover, Ohnishi et al. [17] reported collateral circulation that developed 5–6 weeks after portal obstruction in liver cancer patients. Another study by Ohno et al. [18] experimentally proved that marked collateral circulation to the liver was
likely to develop relatively early (within 72 h). In our department, marked cavernomatous transformation was noted 4 months after the onset of portal thrombosis, suggesting that this cavernomatous transformation is fully established several months after portal obstruction [19]. With retrograde portography (Celis and Villabolos, 1955 [20]), the contrast medium is forcedly injected from the peripheral hepatic vein, passes through the sinusoid and intrahepatic portal vein, and reaches the portal obstructed region when EHO does not have any block between the intrahepatic portal vein and the hepatic vein, so that the obstructed region in the intrahepatic portal vein can be confirmed [21]. Retrograde portography useful for our studies of examination of the intrahepatic obstructed region. The present study examined the changes of intrahepatic portal obstruction in EHO patients after long-term followup using retrograde portography. In 7 patients under long-term follow-up, the region of the EHO and the morphology of collateral circulation to the liver changed, however intrahepatic portal vein obstruction was not extended. As for the reason of the difference, we suggest that the progression of intrahepatic portal obstruction comes to an end at the time of the development of cavenomatous transformation, while extrahepatic portal obstruction progresses because of the occurrence of new thrombus in consequence of portal stasis, and the form of cavenomatous transformation changes. Liver functions are maintained satisfactorily for a longer time in EHO, one of the causal diseases of portal hypertension, than in LC or IPH. The prognosis for EHO is good if recurrent esophageal varices are appropriately treated. Based on these results, EHO patients showed continuous changing of EHOs during long-term follow-ups and had a high recurrence rate of esophageal varices after various treatments due to persistent portal hypertension. Lastly, from the outcome of the present study, we suspect the development of cavernomatous transformation and the absence of the progression of intrahepatic portal vein obstruction result in the good prognosis of EHO.
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
Acknowledgements The authors thank Dr Jo Ariyama and Dr Masafumi Suyama (Department of Gastroenterology, Juntendo University, Tokyo) for their technical support of angiography and the analyses of the findings.
References [1] Kameda H. The diagnostic guide for EHO. Surveys and research on specific diseases for disorders of portal circulation from the Ministry of Health, Labour and Welfare (in Japanese). 1986; 270. [2] Heijboer H, Brandjes DPM, Buller HR, Sturk A, Tencate JV. Deficiencies of coagulation-inhibiting and fibrinolytic proteins in outpatients with deep-vein thrombosis. N Engl J Med 1990;323:1512 –6. [3] Webb LJ, Sherlock S. The aetiology, presentation and natural history of extra-hepatic portal venous obstruction. Q J Med 1979;48:627 –39. [4] Imai S, Yamazaki K. The report of survey for IPH and EHO. Research report in 1986. Surveys and research on specific diseases for disorders of portal circulation from the Ministry of Health, Labour and Welfare (in Japanese). 1986; 19 –26. [5] Brown KM, Kaplan MM, Donowitz M. Extrahepatic portal venous thrombosis: frequent recognition of associated diseases. J Clin Gastroenterol 1985;7:153 – 9. [6] Odievre M, Pige G, Alagille D. Congenital abnormalities associated with extrahepatic portal hypertension. Arch Dis Child 1977;52:383 –5. [7] Ando H, Kaneko K, Ito F, Seo T, Watanabe Y, Ito T. Anatomy and etiology of the extrahepatic portal vein obstruction in children leading to bleeding esophageal varices. J Am Coll Surg 1996;183:543 –7. [8] Thompson EN, Sherlock S. The aetiology of portal vein thrombosis with particular reference to the role of infec-
151
tion and exchange transfusion. Q J Med 1964;33:465 – 80. [9] Boles ET Jr, Wise WE Jr, Birken G. Extrahepatic portal hypertension in children. Long term evaluation. Am J Surg 1986;151:734 – 9. [10] Balfour GW, Stewart TG. Case of enlarged spleen complicated with ascites, both depending upon varicose dilatation and thrombosis of the portal vein. Edinb Med J 1868;14:589 – 98. [11] Klemperer P. Cavernous transformation of the portal vein. Its relation to Banti’s disease. Arch Path 1928;6:353 – 77. [12] Bolt RJ. Diseases of the hepatic blood vessels. In: Bockus HL, editor. Gastroenterology, 3rd ed. Philadelphia: WB Saunders Company, 1976:475 Chapter III. [13] Gibson JB, Richards RL. Cavernous transformation of the portal vein. J Path Bact 1995;LXX:81. [14] Rosch J. Extrahepatic portal obstruction in childhood and its angiographic diagnosis. Am J Roentgenol 1971;112:143. [15] Honna T. Experimental models of cavernomatous transformation. Research report in 1986. Surveys and research on specific diseases for disorders of portal circulation from the Ministry of Health, Labour and Welfare (in Japanese). 1986; 169 – 73. [16] Ibuki Y, Hirasa M, Takakura H, et al. A case of supportive pylephlebitis with development of hepatopetal collaterals. Jpn J Gastroenterol 1985;82:1406 – 11 in Japanese. [17] Ohnishi K, Okuda K, Ohtsuki T, et al. Formation of hilar collaterals or cavernous transformation after portal vein obstruction by hepatocellular carcinoma. Observation in ten patients. Gastroenterology 1984;87:1150 – 3. [18] Ohno Y, Ide T, Kanmatsu T. Clinical features of extrahepatic portal vein thrombosis in Japan. J Hepatobiliary Pancreat Surg 1999;38:45 – 50 in Japanese. [19] Yoshimoto J, Takei M, Kitayama N, et al. A case of extrahepatic portal vein obstruction with portal vein aneurysm. Acta Hepat Jap 1997;38:654 – 9 in Japanese. [20] Celis A, Villabolos ME. Roentgenographic opacity of the hepatic circulation. Am J Roentgenol 1955;74:1089 – 95. [21] Fukasawa M, Sugiura M. Angiographic artalations in IPH and EHO. Nihonigakukan 1988:52 – 7 in Japanese.
www.elsevier.com/locate/ihepcom
Hemodynamics in extrahepatic portal vein obstruction and its changes during long-term follow-ups Akiko Ogawa *, Shigeru Takamori, Kuniaki Kojima, Masaki Fukasawa, Tomoe Beppu, Shunji Futagawa Second Department of Surgery, School of Medicine, Juntendo Uni6ersity, 2 -1 -1 Hongo, Bunkyo-ku, Tokyo 113 -8421, Japan Received 26 November 2001; received in revised form 4 February 2002; accepted 8 February 2002
Abstract Thirty-six patients with extrahepatic portal vein obstruction (EHO) were studied. Twenty-one patients, whose age at onset was under 20 years old, were thought to be primary EHO. Ten of 15 patients over 21 years old, had a history of abdominal diseases, and were thus considered to be secondary EHO; only 5 were primary EHO. EHO was classified into three categories, based on selective celiac-superior mesenteric arteriography, and intraoperative trans-splenic and superior mesenteric venography: the hilar obstruction of portal vein, the portal trunk obstruction, and the portal trunk–splenic vein obstruction. Both primary and secondary EHO cases were found to have extended their obstructed region over time by progressing from hepatic hilum to portal trunk obstruction followed by splenic vein obstruction. On the other hand, based on retrograde portography, EHO was classified into the absence of intrahepatic portal obstruction (portal trunk type) and the intrahepatic obstruction (portal branch type) and the portal branch type tended to predominate. Changes of intra- and extra-hepatic portal obstruction were examined by comparing various angiographies of the 14 patients available for between 1 and 26 years of long-term follow-up. While EHO tended to progress with time, no patient showed progress of a intrahepatic portal vein obstruction. From these results, EHO patients are expected to show continuous progression of EHOs during long-term follow-ups, and to have high recurrence rates of gastrointestinal varices after various treatments due to the persistent portal hypertension. However, effective hepatic flow is likely to be maintained because collateral circulation to the liver develops, and intrahepatic portal vein obstruction does not progress, resulting in a good prognosis for EHO. © 2002 Elsevier Science B.V. All rights reserved. Keywords: Portal hypertension; Cavernomatous transformation; Intrahepatic portal vein obstruction
1. Introduction Extrahepatic portal vein obstruction (EHO) is the third greatest cause of portal hypertension * Corresponding author. Tel.: +81-3-5802-1078; fax: + 813-3818-7589
following hepatic cirrhosis (LC) and idiopathic portal hypertension (IPH). While its prognosis is better than LC and IPH, EHO is characterized by a high recurrence rate of varices after various treatments and by a wide age of onset (from children to adults). Additionally, it is said that there are a wide variety of causes of portal ob-
1386-6346/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 6 - 6 3 4 6 ( 0 2 ) 0 0 0 2 9 - 3
142
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
struction. The present study examined the portal obstructed region using various angiographies, the morphology of cavernomatous transformation and their changes during long-term follow-up in the patients with EHO. 2. Subjects and methods The subjects were 36 EHO patients (23 males and 13 females, 5– 76 years old, average 38.2 years) admitted to our department since September 1979 (Table 1). In the table, we defined the age of onset as the age at which the clinical symptoms developed. EHO was diagnosed by selective celiac and superior mesenteric arteriography at the initial operation for esophagogastric varices that were preoperatively performed in all cases, and by intraoperative trans-splenic and superior mesenteric venography. We defined the site of obstruction in EHO as the defect in portography including intraoperative angiography in the present study, because it was so difficult to detect the site of obstruction accurately with intraoperative findings. As to intraoperative angiography, although trans-superior mesenteric venography was applied before 1985, trans-splenic venography has been performed in all cases excluding the cases with a history of splenectomy since we experienced a complication of portal thrombus after trans-superior mesenteric venography. From the findings of angiographs, EHO was classified into three categories (Fig. 1): (1) the hilar obstruction of portal vein type; the portal trunk opened and only the hilar region of portal vein was obstructed, (2) the portal trunk obstruction type; the entire portal trunk was obstructed and replaced with cavernomatous transformation, and (3) the portal trunk –splenic vein obstruction type; the portal trunk and splenic vein were obstructed. The findings of venographies did not be affected by the methods such as preoperative and postoperative venographies. Angiography also showed that cavernomatous transformation has been developed in all type. The cavernomatous transformations flowed almost evenly into both branches of portal vein and there was no case in which the dominant flow was found into either branch of portal vein.
The morphology of EHO was further classified into two categories by retrograde portography performed by wedging a catheter into the peripheral hepatic vein during hepatic vein catheterization (Fig. 2): (1) the absence of intrahepatic portal obstruction (portal trunk type) (extrahepatic type); intrahepatic portal vein obstruction was limited in the hepatic portal region and communication between right and left hepatic portal veins was observed, and (2) the intrahepatic obstruction (portal branch type) (intrahepatic type); obstruction extended over the intrahepatic portal vein and no communication between the right and left hepatic portal vein was observed. These categories were associated with age of onset, history of abdominal diseases, and duration of symptoms in the 36 subjects. The retrograde portography was carried out with manual injection to prevent a complication of sub-capsule hematoma of the liver. EHOs of No. 27, 28, 35 those complicated with LC or IPH were excluded from the study on retrograde portography because the contrast on the angiography was restricted by presinusoidal or postsinusoidal blocks. On the other hand, because contrast medium was injected into portal vein without much resistance, injected pressure did not affect the examination of obstructed portion in the other EHOs. The retrograde portographies were performed on 32 cases with informed consent excluding emergent cases, cases without agreement to the angiography, and cases with previous operation at other hospital. Various angiographies in during the follow-up periods were also compared with each other and the morphological changes of intra-/extra-hepatic portal vein obstruction and cavernomatous transformation over time were observed in the 14 patients who were available for 1– 26 years of follow-up study after their initial operations. In the follow-up study, we defined the extension of the defect in venography as the progression of EHO.
3. Results The age of onset of the 36 EHO patients ranged from 1 to 64 years with the most common initial symptoms of hematemesis/melena, followed by
Age of onset
1 2 4 4
4 4
4 5 7 7 10 12 14 14 15 15 16
16
16 17 19
25
27 27 32 39
48
Age and sex
11 ß 5à 16 ß 16 ß
23 ß 28 ß
40 ß 40 à 7ß 37 ß 22 à 13 à 14 ß 41 ß 17 à 37 ß 17 ß
21 à
53 ß 17 ß 53 à
33 à
ß ß ß ß
27 28 34 41
50 ß
Case
1 2 3 4
5 6
7 8 9 10 11 12 13 14 15 16 17
18
19 20 21
22
23 24 25 26
27
Table 1 Clinical features of patients with EHO
Esophagogastric varices Melena Abdominal pain Splenomegaly Esophagogastric varices Esophagogastric varices
Abdominal distension Unknown Hematemesis Nasal hemorrhage
Splenomegaly Melena Hematemesis Hematemesis and melena Splenomegaly Hematemesis and melena Splenomegaly Anemia Hematemesis Unknown Hematemesis Pancytopenia Splenomegaly Hematemesis Abdominal pain Splenomegaly Splenomegaly
Clinical findings
Ope for appendicitis (15), for peritonitis (15), for ileus (20)
Portal thrombosis (27) Ope for appendicitis(9) Ope for appendicitis (16)
Exchange transfusion via umbilical vein
Past history of abdominal diseases (age)
2
0 1 2 2
8
37 0 34
5
36 35 0 30 12 1 0 27 2 22 1
19 24
10 3 12 12
Period between onset and angiography (years)
H
S P S S
P
P H H
S
P P H P H H P H P P P
P H
H H H H
Portal phase of arteriography
I
I I I I
I
I E E
I
I I I I I I I – I I I
I I
E I E E
Retrograde portography
Site of obstruction
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151 143
49
52
54
55
55 56
61 62
64
50 à
52 à
59 ß
55 à
58 ß 57 à
65 ß 62 ß
76 à
28
29
30
31
32 33
34 35
36
Esophagogastric varices
Esophagogastric varices Point out at operation Hematemesis By ultrasonography Unknown Esophagogastric varices
Esophagogastric varices Esophagogastric varices
Clinical findings
Chronic pancreatitis (51) Ope for gastric ulcer (46), for choledocholithiasis and liver abscess (60)
Choledocholithiasis (16), cholangitis (55) Cholelithiasis (51)
Ope for cholelithiasis and choledocholithiasis (36), for peritonitis (47)
Cholecystitis (49)
Past history of abdominal diseases (age)
12
4 0
3 1
0
5
0
1
Period between onset and angiography (years)
Ope, operation; H, hilum; P, portal trunk; I, intrahepatic; S, P and splenic vein; E, extrahepatic.
Age of onset
Age and sex
Case
Table 1 Clinical features of patients with EHO
P
P H
P P
H
P
P
P
Portal phase of arteriography
–
I I
– I
–
E
I
I
Retrograde portography
Site of obstruction
144 A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
esophageal/gastric varices and splenomegaly (Table 1). Patients with any past history seemed to be a causal factor of EHO were a case who had undergone neonatal exchange transfusion via the umbilical vein, 5 cases who had cholelithiasis/ choledocholithiasis, and 5 cases who underwent abdominal operation.
3.1. Age of EHO onset and types of intra-/extra-hepatic portal obstruction Table 2 provides a summary of the site of obstruction and the age of onset. More than half
145
(11 of 21, 52.3%) patients whose age of EHO onset was under 20 years had hilar obstruction and only one patient (4.8%) had a portal trunk– splenic vein obstruction. In patients whose age of EHO onset was over 21 years, no more than 4 of 15 patients (26.7%) had hilar obstruction and 3 patients (20.0%) had portal trunk–splenic vein obstruction, showing that the EHO tends to be diffuse. In all cases, cavernomatous transformations were found to extend beyond the portal obstructive region to the liver. From the findings of retrograde portography in 31 patients with EHO the portal trunk type extrahepatic type con-
Fig. 1. Portographies show (a) hilar obstruction of portal vein type, (b) portal trunk obstruction type and (c) portal trunk – splenic vein obstruction type, in which the portal trunk is obstructed in portal phase of superior mesenteric arteriography (c-1) and splenic vein of celiac arteriography (c-2). Arrows show the part of obstruction.
146
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
Fig. 2. Retrograde portgraphies show (a) extrahepatic type (cannulated through left hepatic vein) and (b) intrahepatic type (cannulated through right hepatic vein). Arrows show the site of obstruction.
sist of 5 patients (25.0%) whose age of onset was under 20 years and a patient (11.1%) over 21 years, and the portal branch type intrahepatic type of 15 (75.0%) and 8 (88.9%), showing that the portal branch type tended to increase in the patients whose age at onset was over 21 years old (Table 3).
3.2. History of abdominal diseases and types of intra-/extra-hepatic portal obstruction In the patients whose age at EHO onset was over 21 years, 3 of 5 patients without a history of abdominal diseases had portal trunk obstruction. In patients with a history of abdominal diseases, 3 patients had hilar obstructions (30.0%), 5 had portal trunk obstructions (50.0%), and 2 had portal trunk –splenic vein obstruction (20.0%); showing a slightly lower ratio of hilar obstruction (Table 4). In 12 patients whose age at EHO onset was over 21 years and who underwent retrograde portography, of those with a history of abdominal disease, none had the portal trunk type extrahepatic type and 5 (100%) had the portal branch type intrahepatic type. In patients without a history of abdominal disease, one had portal trunk type (25.0%) extrahepatic type and 3 the portal branch type
intrahepatic type (75.0%). In both cases the numbers of patients with portal branch type intrahepatic type were higher than with portal trunk type extrahepatic type (Table 5).
Table 2 Age of onset with EHO and site of obstruction Site of obstruction
Hepatic hilum Portal trunk Portal trunk and splenic vein
Cases
15 17 4
Age of onset (years) 0–20
Over 21
11 9 1
4 8 3
Table 3 Age of onset with EHO and site of obstruction from the findings of retrograde portography Site of obstruction
Cases
Extrahepatic portal vein 6 Intrahepatic portal vein 23
Age of onset (years) 0–20
Over 21
5 15
1 8
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
147
Table 4 Past history of abdominal diseases with onset age of over 21 years and site of obstruction Site of obstruction
Cases
Abdominal diseases (−)
Abdominal diseases (+)
Hepatic hilum Portal trunk Portal trunk and splenic vein
4 8 3
1 3 1
3 5 2
Table 5 Past history of abdominal diseases with onset age of over 21 years and site of obstruction from the findings of retrograde portography Site of obstruction
Cases
Abdominal diseases (−)
Abdominal diseases (+)
Extrahepatic portal vein Intrahepatic portal vein
1 8
1 3
0 5
Table 6 Period after abdominal disease or operation at onset age with EHO of over 21 years and site of obstruction Site of obstruction
Hepatic hilum Portal trunk Portal trunk and splenic vein
Cases
3 4 2
In the patients with a history of abdominal diseases, 2 of 4 whose period from suffering the disease to angiography was within 2 years had hilar obstruction, 3 of 3 whose period was 3– 10 years had portal trunk obstruction, and 2 of 3 whose period was over 10 years had portal trunk –splenic vein obstruction. The number of patients with portal trunk and portal trunk– splenic vein obstructions tended to increase with time (Table 6).
3.3. Changes of portal obstruction in the patients subjected to long-term follow-up Various angiographies at the initial operation and during the follow-up period were compared with each other to examine the changes of intra-/ extra-hepatic portal obstruction in 14 patients
Period after abdominal disease or operation (years) Below 2
3–10
Over 10
2 1 0
0 3 0
1 0 2
who were available for long-term follow-up ranging from 1 to 26 years. Four cases of hilar obstruction showed an extension to the portal trunk obstruction, and one portal trunk obstruction case changed to the portal trunk–splenic vein obstruction. We present a case, with melena as an initial symptom, that intra-operative venography at the operation of esophageal transection for esophagogastric varices at the age of 5 and the portographies after 12 years and 16 years after operation were available in Fig. 3, and retrograde portographies at the operation and 12 years after operation were also available in Fig. 4. Intra-operative venography showed that the portal trunk opened and only the hilar region of portal vein was obstructed (hilar type). At the 12 years after operation, the portal trunk still opened. At the 16 years after operation the obstruction developed to the portal trunk, and the morphology of caver-
148
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
nomatous transformation also changed. The retrograde portography showed the obstruction was portal branch type intrahepatic type at the operation, and 12 years after operation the obstruction did not developed (Table 7). On the other hand, in the 5 patients who twice underwent retrograde portography, none showed extensions of intrahepatic portal obstructions.
4. Discussion According to the guidelines for the diagnosis of EHO made by Ministry of Health, Labour and Welfare Specified Diseases Study Group for Abnormal Portal Circulation [1], EHO can be divided into two categories: primary EHO, whose causes are unknown, and secondary EHO, whose
Fig. 3. Portographies show (a) hilar obstruction at the operation for esophagogastric varices in the intra-operative venography, (b) 12 years after operation and (c) portal trunk obstruction in the portal phase of superior mesenteric arteriography at 16 years after operation.
Fig. 4. Retrograde portographies show (a) intrahepatic obstruction at the time of operation and (b) 12 years after operation. No progression of the obstruction is found between at the time of operation and 12 years after operation.
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
149
Table 7 Changes in long-term follow-up portography Case
2 3 20 36 7 8 10 13 16 22 34 23 25 26
Age of onset
2 4 17 64 4 5 7 14 15 25 61 27 32 39
Site of obstruction (period; years) Portal phase of arteriography
Retrograde portography
H(12) H(16) P H P (3) H(10) P(12) P P H (1) P P (1) P(15) P P P (7) P(26) P P P (2) P(11) P P S (1) S S (1) S S (3) S
I(12) I H(10) H I(15) I I(22) I
I I (9)
H, hilum; P, portal trunk; I, intrahepatic; S, P and splenic vein; E, extrahepatic.
causes includes hepatic cirrhosis, IPH, tumor, hematologic disorders, cholecystitis/cholangitis, pancreatitis and abdominal operation. Some reports in recent years have noted that inherited or acquired hypercoagulable states such as protein C, protein S, and antithrombin III deficiency may predispose patients to develop portal venous obstruction [2]. As for the age distribution of EHO, Webb and Sherlock [3] reported that there were two peaks in ages of EHO onset in childhood and adulthood. The Specified Diseases Study Group for Abnormal Portal Circulation also reported from national statistics [4] that a clear peak of EHO onset was found in teenagers and another slight peak was found in the forties in Japan. Since most child patients did not have any causal past history resulting in secondary EHO, several authors have suggested an association between EHO and certain congenital anomalies such as ventricular septal defect, atrial septal defect, deformed inferior vena cava and biliary tract anomalies [5] [6]. Recently, Ando et al. [7] suggested the abnormal running of portal vein, an anomaly during development, as a potential cause. A subsequent study reported that the most common etiologic factor in children is infection, accounting for 43–52% of all cases [8], and umbil-
ical sepsis, the most frequent infection during the neonatal period, is present in 10– 26% of children [9]. One patient in this study whose age at EHO onset was under 20 years had a history of exchange transfusion via the umbilical vein. While there was no history of abdominal diseases in patients whose age of onset was under 20 years, 10 of 15 patients whose age at onset was over 21 years (66.7%) had histories of abdominal diseases. Five of these underwent abdominal operations mainly for the treatment of appendicitis and cholecystolithiasis/choledocholithiasis. Most of appendicitis in these patients had progressed to peritonitis. Five cases were likely to progress from LC or IPH. Five of the 15 patients whose age at onset was over 21 years were thought to have primary EHO, but most were suggested to have so-called secondary EHO. The present study examined the morphology of cavernomatous transformation complicated with EHO. Hilar obstruction was slightly predominant in patients whose age at EHO onset was under 20 years, while extensive obstructions such as portal trunk obstruction including the splenic vein was predominant in patients whose age at EHO onset was over 21 years old. As for primary EHO, hilar obstruction was predominant in patients whose
150
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
age at onset was under 20 years. No hilar obstruction was observed in the patients whose age at onset was over 21 years. Since the ratio of portal trunk obstruction became higher as the age at onset progressed, portal obstruction is suggested to progress from hepatic hilum to portal trunk. Obstruction was found in various regions in secondary EHO. In an examination of the regions of the obstruction based on the period since abdominal disease or operation, EHO tended to progress from hilar obstruction to portal trunk obstruction, and further to splenic vein obstruction over time. Thus both primary and secondary EHO tend to extend their obstruction ranges as years of suffering the condition increase. It has been reported that the most common initial symptom of childhood EHO was hematemesis/melena followed by splenomegaly [3] and the same tendency has been observed in our department. Only one patient had hematemesis as a initial symptom in the patients whose age at primary EHO onset was over 21 years. Cavernomatous transformation to the liver has long been reported. Beginning with Balfour and Stewart in 1868 [10] followed by Klemperer in 1928 who reported additional similar cases [11] and named it cavernomatous transformation based on the histopathological morphology of the hepatic portal region. Nowadays, causal factors of cavernomatous transformation are considered as the recanalization of the portal vein due to portal trunk obstruction and dilated collateral circulation to the liver. Previously it was thought that long-trunk portal obstruction caused cavernomatous transformation [12– 14]. However, Honna [15] reported that the gradual obstruction of the portal vein induced by an ameroid constrictor resulted in development of collateral circulation to/from the liver along with portal hypertension in rats 9 weeks after constrictor application. Ibuki et al. [16] also reported that cavernomatous transformation was completely established 3 months after the development of portal thrombosis. Moreover, Ohnishi et al. [17] reported collateral circulation that developed 5–6 weeks after portal obstruction in liver cancer patients. Another study by Ohno et al. [18] experimentally proved that marked collateral circulation to the liver was
likely to develop relatively early (within 72 h). In our department, marked cavernomatous transformation was noted 4 months after the onset of portal thrombosis, suggesting that this cavernomatous transformation is fully established several months after portal obstruction [19]. With retrograde portography (Celis and Villabolos, 1955 [20]), the contrast medium is forcedly injected from the peripheral hepatic vein, passes through the sinusoid and intrahepatic portal vein, and reaches the portal obstructed region when EHO does not have any block between the intrahepatic portal vein and the hepatic vein, so that the obstructed region in the intrahepatic portal vein can be confirmed [21]. Retrograde portography useful for our studies of examination of the intrahepatic obstructed region. The present study examined the changes of intrahepatic portal obstruction in EHO patients after long-term followup using retrograde portography. In 7 patients under long-term follow-up, the region of the EHO and the morphology of collateral circulation to the liver changed, however intrahepatic portal vein obstruction was not extended. As for the reason of the difference, we suggest that the progression of intrahepatic portal obstruction comes to an end at the time of the development of cavenomatous transformation, while extrahepatic portal obstruction progresses because of the occurrence of new thrombus in consequence of portal stasis, and the form of cavenomatous transformation changes. Liver functions are maintained satisfactorily for a longer time in EHO, one of the causal diseases of portal hypertension, than in LC or IPH. The prognosis for EHO is good if recurrent esophageal varices are appropriately treated. Based on these results, EHO patients showed continuous changing of EHOs during long-term follow-ups and had a high recurrence rate of esophageal varices after various treatments due to persistent portal hypertension. Lastly, from the outcome of the present study, we suspect the development of cavernomatous transformation and the absence of the progression of intrahepatic portal vein obstruction result in the good prognosis of EHO.
A. Ogawa et al. / Hepatology Research 24 (2002) 141–151
Acknowledgements The authors thank Dr Jo Ariyama and Dr Masafumi Suyama (Department of Gastroenterology, Juntendo University, Tokyo) for their technical support of angiography and the analyses of the findings.
References [1] Kameda H. The diagnostic guide for EHO. Surveys and research on specific diseases for disorders of portal circulation from the Ministry of Health, Labour and Welfare (in Japanese). 1986; 270. [2] Heijboer H, Brandjes DPM, Buller HR, Sturk A, Tencate JV. Deficiencies of coagulation-inhibiting and fibrinolytic proteins in outpatients with deep-vein thrombosis. N Engl J Med 1990;323:1512 –6. [3] Webb LJ, Sherlock S. The aetiology, presentation and natural history of extra-hepatic portal venous obstruction. Q J Med 1979;48:627 –39. [4] Imai S, Yamazaki K. The report of survey for IPH and EHO. Research report in 1986. Surveys and research on specific diseases for disorders of portal circulation from the Ministry of Health, Labour and Welfare (in Japanese). 1986; 19 –26. [5] Brown KM, Kaplan MM, Donowitz M. Extrahepatic portal venous thrombosis: frequent recognition of associated diseases. J Clin Gastroenterol 1985;7:153 – 9. [6] Odievre M, Pige G, Alagille D. Congenital abnormalities associated with extrahepatic portal hypertension. Arch Dis Child 1977;52:383 –5. [7] Ando H, Kaneko K, Ito F, Seo T, Watanabe Y, Ito T. Anatomy and etiology of the extrahepatic portal vein obstruction in children leading to bleeding esophageal varices. J Am Coll Surg 1996;183:543 –7. [8] Thompson EN, Sherlock S. The aetiology of portal vein thrombosis with particular reference to the role of infec-
151
tion and exchange transfusion. Q J Med 1964;33:465 – 80. [9] Boles ET Jr, Wise WE Jr, Birken G. Extrahepatic portal hypertension in children. Long term evaluation. Am J Surg 1986;151:734 – 9. [10] Balfour GW, Stewart TG. Case of enlarged spleen complicated with ascites, both depending upon varicose dilatation and thrombosis of the portal vein. Edinb Med J 1868;14:589 – 98. [11] Klemperer P. Cavernous transformation of the portal vein. Its relation to Banti’s disease. Arch Path 1928;6:353 – 77. [12] Bolt RJ. Diseases of the hepatic blood vessels. In: Bockus HL, editor. Gastroenterology, 3rd ed. Philadelphia: WB Saunders Company, 1976:475 Chapter III. [13] Gibson JB, Richards RL. Cavernous transformation of the portal vein. J Path Bact 1995;LXX:81. [14] Rosch J. Extrahepatic portal obstruction in childhood and its angiographic diagnosis. Am J Roentgenol 1971;112:143. [15] Honna T. Experimental models of cavernomatous transformation. Research report in 1986. Surveys and research on specific diseases for disorders of portal circulation from the Ministry of Health, Labour and Welfare (in Japanese). 1986; 169 – 73. [16] Ibuki Y, Hirasa M, Takakura H, et al. A case of supportive pylephlebitis with development of hepatopetal collaterals. Jpn J Gastroenterol 1985;82:1406 – 11 in Japanese. [17] Ohnishi K, Okuda K, Ohtsuki T, et al. Formation of hilar collaterals or cavernous transformation after portal vein obstruction by hepatocellular carcinoma. Observation in ten patients. Gastroenterology 1984;87:1150 – 3. [18] Ohno Y, Ide T, Kanmatsu T. Clinical features of extrahepatic portal vein thrombosis in Japan. J Hepatobiliary Pancreat Surg 1999;38:45 – 50 in Japanese. [19] Yoshimoto J, Takei M, Kitayama N, et al. A case of extrahepatic portal vein obstruction with portal vein aneurysm. Acta Hepat Jap 1997;38:654 – 9 in Japanese. [20] Celis A, Villabolos ME. Roentgenographic opacity of the hepatic circulation. Am J Roentgenol 1955;74:1089 – 95. [21] Fukasawa M, Sugiura M. Angiographic artalations in IPH and EHO. Nihonigakukan 1988:52 – 7 in Japanese.