Vascular and Coagulation Disorders of the Liver

The liver has three vascular systems-the portal vein, the hepatic vein, and the hepatic artery-but the hepatic artery seldom causes problems, unlike other arteries in other parts of the body. The arterial system of the body develops obliterative changes with morbidity as the age advances. Although the hepatic artery, splenic artery, and mesenteric arteries originate from the same aorta in the abdomen, the hepatic artery rarely develops insufficiency. Atherosclerotic changes in the mesenteric artery cause malperfusion that leads to abdominal angina; aneurysm of the abdominal aorta is relatively common in the aged. The hepatic artery seems to be spared of such changes. Calcification of the abdominal aorta and the splenic artery is particularly common and severe in patients with chronic renal failure because of hyperparathyroidism, but the hepatic artery seldom calcifies. The liver is vital for life, and certain ill-defined mechanisms are operative in preventing calcific and atherosclerotic changes of the hepatic artery, perhaps serving a teleologic purpose. By contrast, the venous systems of the liver are amenable to changes that cause circulatory disturbances, such as thrombosis, varices, and collateral veins. Thus, this issue of Seminars in Liver Disease deals only with venous systems of the liver.

Budd-Chiari syndrome was primarily described in patients with hepatic vein thrombosis within the liver induced by infection, but subsequent reports on similar clinical symptoms by other causes obscured the distinction between primary hepatic vein thrombosis and thrombosis of the inferior vena cava (IVC) affecting orifices of hepatic veins. The clinical features of classical hepatic vein thrombosis (discussed by Valla in this issue) and primary thrombosis of the hepatic portion of the IVC affecting hepatic vein orifices (described by Okuda) are quite different; the former is clinically more severe, sometimes requiring liver transplantation. Primary IVC thrombosis, which preferentially occurs in the hepatic portion of the IVC and eventually turns into a membranous obstruction has milder onset, but recurs frequently, and slowly progresses into congestive cirrhosis, which may be complicated by hepatocellular carcinoma. In southern Africa, among the Bantus nearly half the patients with membranous IVC obstruction develop hepatocellular carcinoma, whereas this complication is extremely rare in primary hepatic vein thrombosis. A number of thrombogenic or hypercoagulable disorders are known to cause hepatic vein thrombosis, and Valla discusses them with emphasis on myeloproliferative disorders. Earlier Ludwig, and more recently Okuda, emphasized the need for separating hepatic vein thrombosis and IVC thrombosis and abandoning the term Budd Chiari syndrome; the suggested term for IVC thrombosis is obliterative hepatocavopathy. In France, etiologic or thrombogenic disorders are found in nearly all cases of primary hepatic vein thrombosis. The same does not apply to IVC thrombosis, which is frequently idiopathic with no demonstrable hypercoagulability. In Nepal, where IVC thrombosis or obliterative hepatocavopathy is endemic and is one of the most common liver diseases, patients frequently have bacterial infection or bacteremia. Such an etiology also may apply to some of the sporadic cases of obliterative hepatocavopathy elsewhere. The membranous obstruction of IVC was once thought by many to represent a congenital vascular malformation or a coarctation comparable with aortic coarctation. This disease mostly occurs in adults, and the congenital theory has recently been completely denied. The membranous obstruction is not always a membrane, and could be much thicker or thinner; it is unequivocally a sequela to IVC thrombosis.

Veno-occlusive disease (VOD) is an ailment that affects central veins, the most peripheral hepatic veins. It was first described in a family in Cape Town, South Africa, who coharvested wheat and the seed of Senecio, a weed. Eating bread contaminated with Senecio produced a cirrhosis-like disease with ascites. Subsequently, a nonportal type of cirrhosis was reported among the inhabitants of Jamaica who drank herbal teas containing pyrrolizidine. A number of small outbreaks of a severe liver disease were observed in Afghan villages where toxic weed contaminated grain, and dogs also developed ascites and died. More recently, it was found that immunosuppressants, notably azathioprine, and chemotherapeutic agents also cause VOD; many VOD cases associated with kidney and liver transplantation have been documented. Histopathologically, it was primarily thought that VOD represented an obliterative disease of hepatic venules. Recent studies suggest, however, that the primary site of toxic injury is sinusoidal endothelium followed by a series of biologic processes that lead to circulatory disturbances compromising centrilobular hepatocytes and cause fibrosis and obstruction to hepatic blood flow. It was also found that hepatic venule involvement is not essential to the development of clinical symptoms. McDonald has proposed that the disease be renamed sinusoidal obstruction syndrome (SOS). In this issue, McDonald and associates describe in detail the histogenesis of SOS and clinical manifestations in patients who received high-dose myeloablative therapy in preparation for hematopoietic stem cell transplantation.

Among various noncirrhotic disorders that cause portal hypertension, idiopathic portal hypertension (IPH) in Japan and noncirrhotic portal fibrosis (NCPF) in India are more common than other portal hypertensive diseases. IPH, first described by Banti as a spleen disorder toward the end of the 19th century in Italy, is the disease most extensively studied in Japan. Okudaira et al discuss the results of the study of IPH conducted in Japan by a research group supported by the government. The nation-wide survey made by this group indicates that currently less than 20 new cases occur per year in the country. They describe the vascular anatomy characteristic of IPH, and refute the theory that IPH merely represents undiagnosed intrahepatic portal vein thrombosis. In other words, IPH is a distinct disease entity, although the exact etiology is unknown.

Before 1970, more than 30% of patients with portal hypertension were noncirrhotic in Japan, mostly IPH. A drastic reduction in IPH incidence has occurred in the past 30 years in this country, and IPH is no longer a common disease. In India, NCPF and extrahepatic portal vein obstruction (EHPVO) constituted more than 50% of portal hypertensive patients when Ramalingaswami first described NCPF among autopsies in 1962, and the incidence or epidemiology of NCPF has not changed since. A major difference between the two countries is the incidence of bacterial infection, which has become uncommon due to improved living in Japan; the same has not occurred in India. Clearly, bacterial infection is etiologically involved in a manner not well defined.

In India, both NCPF and EHPVO are common. Sarin and Agarwal describe in this issue EHPVO seen in India, which is perhaps closely associated with bacterial infection. In adults, portal biliopathy as a result of collateral paracholecystic veins is common, but it is rare in Japanese EHPVO patients, an interesting difference. There are cases in which thrombosis exists both extrahepatically and intrahepatically, and a clear distinction of NCPF and EHPVO is difficult in such cases. In the absence of thrombotic changes in the mesenteric vein, more etiologic study is needed for differentiating NCPF and EHPVO. Although the etiology of IPH and NCPF is not known at the moment, recent studies in Japan suggest immunologic processes as a cause of IPH. Experimentally, introduction of killed colon bacilli into the portal vein rapidly produces marked portal fibrosis and portal hypertension. Data are lacking on the reaction of the portal vein wall and portal tract to stimuli within the portal circulation, and more animal models are required for the elucidation of etiology of IPH and NCPF.

There has been surprisingly limited information on the vascular pathology of varices. Arakawa et al describe in this issue the pathomorphology of esophageal and gastric varices studied in autopsy material using specific techniques that facilitate the gross anatomic study of the vasculature in relation to mural tissue layers, namely the lamina propria mucosae, submucosa, muscularis, and serosa. They emphasize the thin parallel veins that coalesce at an acute angle with submucosal varicose veins in the palisade zone of the esophagus. They call this particular location the ``critical area,'' where rupture commonly occurs because of a sudden change in the course of the veins. The basic difference between esophageal and gastric varices is their different locations, the former forming in the lamina propria mucosae and the latter in the submucosa. In this article, the gross vascular anatomy of fundic varices is detailed in relation to the feeding veins. Because fundic varices mostly consist of a single large collateral vein involving the splenic and left renal veins, bleeding is massive and difficult to manage. Various therapeutic approaches are also discussed for esophageal and fundic varices in comparison. A new technique developed by Japanese endoscopists called balloon occluded retrograde transvenous obliteration is briefly touched on. In this method, the catheter is inserted into the fundic varices through the left renal vein and the varices embolized.

The clotting process is markedly affected by various liver disorders because of the decreased synthesis of clotting and inhibitor factors, decreased clearance of activated factors, quantitative and qualitative platelet defects, hyperfibrinolysis, and accelerated intravascular coagulation. Amitrano and associates discuss these aspects in relation to the liver, and the peculiar coagulation disorders in patients with acute fatty liver of pregnancy and in patients undergoing liver transplantation. They also discuss the recent evidence that sepsis impairs hemostasis in patients with cirrhosis bleeding from varices.

There are other aspects that are not covered in this issue, such as congenital vascular anomalies, sequelae to compromised portal perfusion, intrahepatic and extrahepatic portal vein aneurysm, and so forth. I hope that these aspects will be taken up in a future issue of Seminars in Liver Disease.

Kunio Okuda, M.D., Ph.D.

Guest Editor