The preferable treatment for cirrhotic portal vein thrombosis: Anticoagulation or transjugular intrahepatic portosystemic shunt?

To the Editor:

We read with great interest the American Association for the Study of Liver Diseases practice guideline1 showing the current central status of anticoagulation for portal vein thrombosis (PVT), although anticoagulation should never be stereotyped for all patients with PVT. It is very important for clinicians to distinguish between noncirrhotic and cirrhotic PVT because of its association with therapeutic strategy selection.

Nonmalignant and noncirrhotic PVT is mainly caused by inherited and acquired prothrombotic disorders, except for some local factors, so the treatment should be focused on correction of these disorders and not on the creation of a hepatic parenchymal shunt and thrombectomy, which may be just a temporary rescue. According to our data for the period of December 2001 to September 2008, the outcome of noncirrhotic PVT by transjugular intrahepatic portosystemic shunt (TIPS) was unsatisfactory with respect to the technical success rate (9/23) and follow-up in patients with successful recanalization. In contrast, the outcome by anticoagulation therapy was inspiring, as many recent reports have shown.2, 3

The primary cause of cirrhotic PVT is portal flow reduction and its secondary hemostasis, even if other factors play a role, such as the decline of coagulation inhibitors synthesized by the liver and inherited coagulation abnormalities. Variceal rebleeding is one of the most common clinical presentations in patients with decompensated cirrhosis and PVT, at least in our aforementioned data (52/61). Therefore, the goal of treatment in theory is to deal with the portal hypertension and smooth portal vein. Benefits lie not only in a successful recanalization rate (43/61) but also in postoperative follow-up in our experience (Table 1). It seems that TIPS with thrombectomy should be adopted more widely than anticoagulation therapy (Fig. 1), which is still challenged by the unresolved issue4 of whether to aggravate the risk of bleeding or not, especially for patients with thrombocytopenia and large varices.

Table 1. Preliminary Outcomes of 43 Patients with Successful Recanalization by TIPS in Our Center

Main Item	Number of Patients
Abbreviations: OLT, orthotopic liver transplantation; SPSS, surgical portosystemic shunt; TIPS, transjugular intrahepatic portosystemic shunt. * Patients who died of TIPS-related complications (particularly shunt dysfunction).
Hepatic capsule perforation	2
Shunt dysfunction	17
<1 year	8
>1 year	9
Hepatic encephalopathy	10
<1 year	7
>1 year	3
Hepatic myelopathy	4
Hepatocellular carcinoma	4
OLT or SPSS	2
Death*	5
<1 month	1
<6 months	3
>6 months	1

Figure 1. Comparison of TIPS and anticoagulation. The cube represents a patient with cirrhosis and portal vein thrombosis in whom variceal rebleeding is a major presentation. Abbreviation: TIPS, transjugular intrahepatic portosystemic shunt.

In addition, another point has to be emphasized: PVT with cirrhosis should never be regarded as cirrhotic PVT, and PVT without cirrhosis should never be regarded as noncirrhotic PVT. From our perspective if any one of the following conditions was identified, this further excluded cirrhotic PVT, even if cirrhosis had been diagnosed:

• 1
  Definite primary thrombosis of the hepatic vessels was diagnosed. For example, Budd-Chiari syndrome had been recognized and treated a few years before the diagnosis of cirrhosis and PVT.
• 2
  Other local factors clearly existed, such as pancreatitis or splenectomy.
• 3
  Cirrhosis was in a compensated condition. In reverse, PVT with well-compensated cirrhosis should not be excluded from noncirrhotic PVT.

The effect can be better only if we aim at the major etiology more accurately and choose a corresponding treatment. It is time to establish the feasibility and safety of TIPS for cirrhotic PVT and thereby design prospective studies, although we have to acknowledge the technical difficulty of TIPS with thrombectomy for PVT: only 28 cases have been studied retrospectively in the largest series5 since the first report.6

Xingshun Qi*, Guohong Han*, Daiming Fan*, * Xijing Hospital of Digestive Diseases, Fourth Military Medical University, Xi'an, China.