|Home | About | Journals | Submit | Contact Us | Français|
The incidence of portal vein thrombosis was examined in 885 patients who received orthotopic liver transplantations for various end-stage liver diseases between 1989 and 1990. The thrombosis was classified into four grades. Grade 1 was thrombosis of intrahepatic portal vein branches, grade 2 was thrombosis of the right or left portal branch or at the bifurcation, grade 3 was partial obstruction of the portal vein trunk, and grade 4 was complete obstruction of the portal vein trunk. Among the 849 patients without previous portosystemic shunt, 14 patients (1.6%) had grade 1, 27 patients (3.2%) had grade 2, 27 patients (3.2%) had grade 3 and 49 patients (5.8%) had grade 4 portal vein thrombosis. The incidence of portal vein thrombosis was highest (34.8%) in the patients with hepatic malignancy in the cirrhotic liver, followed by those with Budd-Chiari syndrome (22.2%) and postnecrotic cirrhosis of various causes (15.7%). The patients with encephalopathy, ascites, variceal bleeding, previous splenectomy and small liver had significantly higher incidences of portal vein thrombosis than the others. The total incidence of portal vein thrombosis among the 36 patients with previous portosystemic shunt was 38.9%, which was significantly higher than that (13.8%) of those without shunt.
Thrombosis of the portal vein has been a formidable challenge to orthotopic liver transplantation (1–5). The incidence of portal vein thrombosis was examined in the orthotopic liver transplant recipients with various end-stage liver diseases, and its relationship to several clinical features was studied.
During the 2-yr period between January 1, 1989, and December 31, 1990, 885 patients (522 males and 363 females) received the first orthotopic liver transplantation at the University Health Center of Pittsburgh. Their medical records were carefully reviewed. The ages ranged from 1 mo to 74 yr (mean ± S.D. = 41.2 ± 18.8 yr).
The liver diseases of the patients are listed in Table 1. Portal vein thrombosis was determined by operative findings and pathological examinations of the excised livers at the time of liver transplantation. It was classified into the following four grades: (a) grade 1 = a partial (greater than 50% in diameter) or total thromboembolic and/or sclerotic obliteration of the intrahepatic (segmental) portal vein branches; (b) grade 2 = a partial (greater than 50% in diameter) or total thromboembolic and/or sclerotic obliteration of the right or left portal vein branch or near the bifurcation of the portal vein trunk in which a standard end-to-end portal vein anastomosis was feasible; (c) grade 3 = a partial (greater than 50% in diameter) thromboembolic and/or sclerotic obliteration of the portal vein trunk in which an end-to-end portal vein anastomosis with or without a vein graft was feasible at the junction of the splenic and the superior mesenteric veins of the recipient; and (d) grade 4 = a complete or near complete (more than 90%) obliteration of the portal vein trunk in which an end-to-side portal-superior mesenteric vein anastomosis with a vein graft (a jump graft) or an end-to-end anastomosis at the junction of the splenic and superior mesenteric veins after thromboembolectomy of the veins was created with a vein graft (an interpositional graft).
The incidence of portal vein thrombosis in various liver diseases and its relationship to liver weight and the history of ascites, encephalopathy, spontaneous bacterial peritonitis, gastrointestinal bleeding, endoscopic sclerotherapy for esophageal varices, splenectomy and portosystemic venous shunt were examined. Statistical comparisons between the groups were made using the χ2 test and Student t test. The difference was considered significant when the p value was less than 0.05.
A total of 849 patients had not had the portosystemic shunt before transplantation. A total of 117 (13.8%) of the 849 patients had some degree of portal vein thrombosis. A total of 49 patients (5.8%) had complete obstruction (grade 4), and 27 patients (3.2%) had partial obstruction (grade 3) of the portal vein trunk. Thus a total of 76 patients (9.0%) had surgically significant portal vein thrombosis for orthotopic liver transplantation. The remaining 41 patients (4.8%) had portal vein thrombosis peripheral to the portal vein trunk (grade 1 and grade 2) (Table 1).
High incidences of portal vein thrombosis were observed in the patients with primary hepatic malignancy in the cirrhotic liver (34.8%), those with Budd-Chiari syndrome (22.2%) and those with postnecrotic cirrhosis (15.7%). Portal vein thrombosis in patients with PBC (7.9%), sclerosing cholangitis (3.6%), liver-based inborn errors of metabolism (5.0%) and biliary atresia (4.8%) were significantly less frequent than in those patients with postnecrotic cirrhosis (p < 0.05). The incidences of portal thrombosis were similar (14.3% to 18.1%) among various types of postnecrotic cirrhosis (Table 1).
The portal vein thrombosis was actually caused by tumor emboli in 25 (83%) of the 30 cirrhotic patients with hepatic malignancy and in all of the four noncirrhotic patients with hepatic malignancy.
A total of 36 patients had the portosystemic venous shunt before transplantation. Twenty patients had postnecrotic cirrhosis, five had PBC, four had primary hepatic malignancy with associated cirrhosis and seven had other liver diseases. Because of the small number of cases, the incidences of portal vein thrombosis were examined on the basis of types of the shunt rather than types of liver diseases.
Thirteen patients had portacaval shunts, 7 had mesocaval shunts, 6 had proximal splenorenal shunts and 10 had distal splenorenal shunts. Eleven of the 36 shunts were occluded (4 portacaval, 2 mesocaval, 4 proximal splenorenal and 1 distal splenorenal).
The incidences and the grades of portal vein thrombosis were stratified by the type of shunt as shown in Table 2. Ten (27.8%) of the 36 patients had complete obstruction (grade 4), and 3 patients (8.3%) had partial obstruction (grade 3) of the portal vein.
The relationship between the liver weight and the portal vein thrombosis was examined among the patients with four of the most common liver diseases (postnecrotic cirrhosis, PBC, sclerosing cholangitis and biliary atresia) who had not had the portosystemic shunt (Table 3). The liver weights of patients with postnecrotic cirrhosis were significantly (p < 0.0001) lighter than those with PBC, sclerosing cholangitis and biliary atresia, and the incidence of portal vein thrombosis of the former was significantly (p < 0.02) higher than those of the latter. In the postnecrotic cirrhosis, the liver weights of patients with portal vein thrombosis were significantly (p < 0.02) lighter than those without thrombosis.
The relationship between the portal vein thrombosis and the clinical features (encephalopathy, ascites, spontaneous bacterial peritonitis, gastrointestinal bleeding, endoscopic sclerotherapy for esophageal varices and splenectomy) was examined in the patients without portosystemic shunt (Table 4). The patients with chronic encephalopathy, refractory ascites, gastrointestinal bleeding or splenectomy had a significantly higher incidence of portal vein thrombosis than those without them. The endoscopic sclerotherapy and the spontaneous bacterial peritonitis could not be statistically related to the portal vein thrombosis.
The incidences of portal vein thrombosis in the literature range from 0.6% to 21%, as shown in Table 5 (6–13; Monarca, et al. Gastroenterology 1986;90:509, Correspondence). The reasons for this wide variation are mainly because of the differences in the methods of diagnosis, the liver diseases and the stages of cirrhosis. The lowest incidence of portal vein thrombosis in cirrhosis was 0.6% by Okuda et al. (12), in which the diagnosis was made by portography. In their report the incidence of portal vein thrombosis among the patients with Child C cirrhosis was 1.3% and that of patients after splenectomy was 22.2%. The highest incidence of 21% was reported by Sarfeh (11). In this report all of the 86 patients had had bleedings from esophageal varices and underwent portal decompression surgery after portography. The degree of portal vein thrombosis reported in the literature (6–13; Monarca, et al. Gastroenterology 1986;90;509, Correspondence) appears to be mostly complete (grade 4) or partial (grade 3) obstruction. Including mural thrombosis, Chang and McFadzean (14) reported the incidence of 64.1% (41 of 64) at autopsy.
Complete obstruction of the portal vein was once considered as a contraindication for orthotopic liver transplantation (1). Recent technical advances (2–5) have made orthotopic liver transplantation possible even for the patient with complete portal vein thrombosis. In fact, those patients have not been excluded from orthotopic liver transplantation for the last several years if one of the major tributaries of the portal system has been suitable for venous allografting to supply the portal blood to the transplant liver. Thus the incidence of portal vein thrombosis reported in this study closely represents its actual incidence among the patients with end-stage liver disease who require liver replacement. We also note the high incidence (27.8%) of complete portal vein thrombosis among the patients with previous portosystemic shunt and the high incidence (85.3%) of tumor embolus as a cause of portal vein obstruction among the patients with hepatic malignancy.
Supported by research grants from the Veterans Administration and Project Grant No. DK 29961 from the National Institutes of Health, Bethesda, Maryland.