Liver cirrhosis represents the end stage of chronic diseases of the liver and is associated with life-threatening complications [1, 2]. The natural course of cirrhosis is largely affected by various pathologies, such as variceal bleeding, ascites, and infection [3, 4]. The major predictors of survival of patients with liver cirrhosis are Child-Pugh score; model for end-stage liver disease score; and various biochemical parameters, such as serum bilirubin, albumin, prothrombin time or international normalized ratio, creatinine, as well as encephalopathy and ascites [4, 5]. Recent evidence suggests the association between portal vein thrombosis (PVT) and survival of patients with liver cirrhosis [6], although the data are inconclusive.

Based on the increase in the use of technically-advanced noninvasive liver imaging modalities, PVT has been increasingly identified in patients with cirrhosis, with the estimated current prevalence of PVT in patients with cirrhosis of 0.6 to 26% [3, 7, 8].

The value of anticoagulation in the treatment of PVT in patients with cirrhosis remains controversial [9]. Cirrhosis is associated with bleeding diathesis based on the following factors: prolonged bleeding time, thrombocytopenia associated with hypersplenism, increased prothrombin time/international normalized ratio, reduced synthesis of coagulation factors, and secondary hyperfibrinolysis [10, 11]. More importantly, nearly half of patients with cirrhosis are diagnosed with gastroesophageal varices that can result in life-threatening bleeding events [12]. Taken together, these detrimental conditions greatly limit the use of anticoagulants for patients with cirrhosis. However, recent evidence indicates that both pro- and anticoagulation factors are concomitantly reduced in patients with cirrhosis [13], thereby maintaining a balance in the coagulation system [13–15]. Unfortunately, bleeding risk cannot be accurately assessed in patients with chronic liver disease by globally used coagulation tests [16, 17]. In addition, the occurrence of bleeding in patients with cirrhosis is not primarily dependent on hemostatic abnormalities, but on the severity of portal pressure, endothelial dysfunction, and bacterial infection [14]. Accordingly, the use of anticoagulants for patients with cirrhosis and PVT may be theoretically justified. However, convincing clinicians to prescribe anticoagulants to cirrhotic patients, especially those with decompensated cirrhosis, remains difficult.

Danaparoid sodium (DS) is a glycosaminoglucoronan derived from the same starting material, porcine intestinal mucosa, as unfractured heparin and low-molecular-weight heparins (LMWHs), but its extraction procedure excludes heparin and heparin fragments [18]. Danaparoid is a low-molecular-weight heparinoid consisting of heparin sulfate (84%), dermatan sulfate (12%), and chondroitin sulfate (4%). The mean mass of its components is approximately 6000 Da [18]. Its antithrombotic activity has been well established. Danaparoid catalyzes inactivation of factors Xa (FXa) and thrombin. Like most LMWHs, danaparoid exerts a stronger catalytic effect on inactivation of FXa by antithrombin (AT)-III, than on inactivation of thrombin by AT-III [19].

Anticoagulation therapy is definitely the most effective way to achieve recanalization of the portal vein, thereby improving the prognosis of patients with PVT. To our knowledge, the optimal management of PVT in individuals with cirrhosis has not yet been addressed by any consensus publication or practice guideline. In this study, we examined the response to treatment, recurrence rate and prognosis of PVT in a group of Japanese cirrhotic patients.

The results of this study showed that anticoagulation with DS is safe and effective treatment, significantly reducing the risk of progression of PVT and liver decompensation. In our study, the response rate (CR + PR) was higher than 75% in all the groups of patients tested in this study. These results are to a large extent similar to those reported by Naeshiro et al. [20], who reported response rate of 77% (CR 15%, PR 62%) after 2-week treatment with DS. Interestingly, the treatment was not associated with severe adverse events, such as gastrointestinal bleeding, thrombocytopenia, or worsening of liver dysfunction (Fig. 4).

Although no side effects were encountered, PVT relapse occurred in about third of the patients, though a second course of DS treatment was efficacious in 54% of the patients. In this regard, the second course of DS treatment was followed by warfarin for maintenance therapy. It is important to keep the international normalized ratio (INR) at less than 2 during warfarin use. The prognosis of patients with unresolved PVT remains poor, including death from hepatic failure.

The treatment goals in PVT include the reversal or prevention of progression of thrombosis in the portal venous system and prevention/treatment of complications. Management decisions must be tailored to the individual patient and should be based on the experience of the attending specialist, since there are only a few randomized controlled trials and no standardized treatment protocols are currently available. The current guidelines of the American Association for the Study of Liver Disease [9] recommend that all patients with acute PVT be anticoagulated for at least 3 months, starting with low-molecular-weight heparin (LMWH), followed by an oral anticoagulant. Long-term anticoagulation is recommended for patients with permanent risk factors or with distal extension of the thrombus into the mesenteric veins. In the setting of chronic PVT, patients should be screened for varices and receive appropriate prophylaxis, with long-term anticoagulation therapy being one consideration for patients with risk factors for thrombosis. At present, no standard recommendations exist for patients with cirrhosis.

DS is devoid of heparin or heparin fragments. Its antithrombotic activity has been well established. The exact antithrombotic mechanism of DS is unclear, but is thought to involve a complex interaction between its two major components. Laboratory monitoring is usually not necessary and bleeding enhancement by DS is minimal. However, patients with serum creatinine above 2 mg/dL should be monitored carefully. There is no antidote for DS, and protamine does not reverse its anticoagulant effect. DS is contraindicated in patients with severe hemorrhagic diathesis; active major bleeding; hypersensitivity to danaparoid, sulfates, or pork products; and those with positive in vitro test for antiplatelet antibodies in the presence of danaparoid.

Several studies have investigated the clinical value of warfarin (a vitamin K antagonist) in the treatment of PVT in cirrhotic patients. The rate of PV recanalization in warfarin-treated patients with cirrhosis is about 40% [21]. Orally administered warfarin is more acceptable to patients; however, treatment with warfarin is particularly difficult in patients with cirrhosis, primarily because monitoring of anticoagulation is complex in this particular situation. Notably, the results of assessments based on the INR in patients with liver disease often overestimate bleeding risk, because this index is determined in plasma samples from patients taking warfarin [22]. The INR has only been validated in individuals with normal liver function on stable anticoagulation. In this regard, 29% variation in mean INR values was reported in a study of patients with cirrhosis treated with one of three different thromboplastin reagents [23]. Further studies are needed to determine whether the target INR value ranging between 2 and 3 is adequate in individuals with abnormal INR values before anticoagulation therapy.

Administration of AT-III to patients with cirrhosis might be efficacious in the prevention of PVT. Kawanaka et al. [24] demonstrated that low and decreasing AT-III activity was associated with the development of PVT in patients with cirrhosis who have undergone splenectomy, and that treatment with AT-III concentrate would probably prevent the development of PVT in these patients.

Previous studies showed that the rate of portal vein recanalization was significantly higher while the rate of thrombus progression was significantly lower in the anticoagulation group compared with the non-anticoagulation group. These results suggest that anticoagulation, rather than “wait-and-see” strategy, should be actively employed to maximize the recanalization of thrombosed portal veins in liver cirrhosis. However, this recommendation should be cautiously applied for the following reasons. First, only a small number of studies were included in the two comparative analyses, and none of them was nonrandomized. Second, the role of spontaneous portal vein recanalization should be considered always [9, 21]. To avoid overtreatment, future studies of cirrhotic patients with PVT who benefit most from anticoagulation are warranted.

The timing and duration of follow-up are also controversial. We tend to see patients either in the surgical or specialized coagulation outpatient clinics every 3 months for at least 1 year. Depending on the location and the extent of thrombosis, CECT should be used regularly to assess the vessel patency.

The strength of our study relative to previous studies is that we used DS for 4 weeks rather than for only 2 weeks. However, our study has three important limitations. First, we used DS to treat PVT in patients with >70% stenosis. Treatment adaptation depends on the institution. Second, the study included only a small number of patients who received warfarin as maintenance therapy. We hesitate to use warfarin in patients with cirrhosis. Many novel oral anticoagulants have been approved for such patients, but since warfarin is an antagonist, it can be used relatively safely. Third, we did not examine the long-term outcome of patients after treatment and identified the factors that affect survival.

Warfarin following danaparoid sodium for the treatment of PVT in patients with liver cirrhosis was safe and effective. An early diagnosis of PVT along with the evaluation of the volume of PVT on CT and an early intervention would contribute to the higher efficacy of the treatment. Thus, we recommend anticoagulation for the management of PVT in liver cirrhosis. Prevention of PVT or successful recanalization of a previously thrombosed portal vein can potentially improve survival of such patients.