EASL Clinical Practice Guidelines

Acute portal vein thrombosis (non-cirrhotic, non-malignant)

Definition and scope

Acute PVT is defined as a recent formation of a thrombus within the portal vein and/or right or left branches. The thrombus may extend into the mesenteric or splenic veins; occlusion may be complete or partial. We will limit the discussion to acute PVT occurring in the absence of malignancy and cirrhosis [[54], [75]]. Acute PVT may also occur in patients with long-standing obstruction of portions of the portal venous system [76].



According to prospective [3] and retrospective studies [[40], [77], [78]], acute abdominal pain is present in 90% of acute PVT patients. A systemic inflammatory response syndrome is present in 85% of patients diagnosed with acute PVT which contrasts with local or systemic infection being present in only 20% of these patients. A significant number of patients only have mild non-specific symptoms so that the diagnosis is overlooked and PVT is recognised only at the stage of cavernomatous transformation. Liver tests generally show no, or only mildly and transient abnormalities. Ascites is present in 50% of patients; in most patients only visible on imaging [3]. Due to improved awareness and availability of sensitive non-invasive imaging, diagnosis of portal venous obstruction is now made in 50 to 70% of cases at the stage of acute PVT [[76], [78]].

Course and outcome

Intestinal infarction is the most concerning immediate complication of acute porto-mesenteric vein thrombosis, with a related mortality of up to 60%. Extensive bowel resection may be necessary with a risk of short bowel syndrome [[79], [80], [81], [82]]. The incidence of intestinal infarction has currently declined to 2–20% in patients treated with anticoagulation [3]. In patients not receiving anticoagulation therapy, spontaneous recanalisation of symptomatic PVT appears to be exceptional [83].

Recognising venous mesenteric infarction is difficult as clinical, biological and radiological manifestations are non-specific. Persisting severe abdominal pain despite adequate anticoagulation, organ failure (shock, renal failure, metabolic acidosis, elevated arterial lactates), massive ascites and rectal bleeding all appear to be suggestive of infarction [[79], [80], [81], [82]]. In a recent study, diabetes was the only factor independently associated with intestinal resection [84].


Doppler ultrasound is usually the first imaging procedure performed in the context of abdominal pain. It may detect an absence of flow within the portal vein. The presence of a hyperechoic thrombus in the portal lumen may be lacking [[43], [85]]. Doppler ultrasound, and MR have a lower sensitivity than CT imaging. Doppler ultrasound is dependent on the expertise and awareness of the operator [[43], [85]]. Diagnosis and extension of acute portal venous obstruction should be confirmed by contrast enhanced CT and/or MR imaging. Acquiring images at the correct time (portal phase) is mandatory in order to prevent pitfalls. Images acquired during the late arterial phase are not optimal for the diagnosis of PVT. Furthermore, in cases of low portal vein flow, a delayed arrival of contrast to the portal vein could be seen on CT, giving the appearance of a filling defect resulting in a false positive diagnosis of thrombosis [86].

Portal phase CT scan shows the absence of visible lumen corresponding to the portal vein clot; CT scan provides additional information regarding the extent of the thrombus to the mesenteric veins and arches, the presence of a local factor, or of congestion and ischemia of the bowel. Distal thrombosis (occlusion of second order radicals of superior mesenteric vein), anomalies of the bowel (homogeneous or heterogeneous hypoattenuating or hyperattenuating wall thickening, dilatation, abnormal or absent wall enhancement) or of the mesentery, mesenteric stranding, large ascites, pneumatosis, and portal venous gas are more frequently observed in patients who will need intestinal resection [84].

Studies addressing the duration of PVT are scarce. A recent thrombus can be defined as a thrombus occurring in the setting of abdominal pain and or systemic inflammatory response syndrome. A spontaneous hyperdense clot in the portal vein lumen on a non-enhanced CT scan may suggest that the thrombus dates back to less than 30 days after onset of symptoms [43]. Absence of portal cavernoma is also of help, although cavernoma may not develop in unilateral portal branch obstruction. A cavernoma may be identified as early as 15 to 30 days after the apparent onset of abdominal pain [43]. Furthermore, acute thrombosis may superimpose on a long-standing cavernoma.

Underlying prothrombotic disorders and local factors are common in adults. These disorders constitute major determinants of outcome, and may require specific therapy (see section 1). In children, aetiological investigations have been negative or only show common weak prothrombotic conditions [87].


The aim of therapy for acute PVT is; i) to prevent the extension of thrombosis to mesenteric veins and thereby, mesenteric venous infarction; and ii) to achieve portal vein recanalisation (Fig. 2) [[43], [85]].

Fig. 2
Proposed algorithm for the management of acute portal vein thrombosis.

View Large Image | View Hi-Res Image | Download PowerPoint Slide


In a recent prospective study, thrombus extension was prevented in all patients who had early initiation of anticoagulation therapy [3]. Only 2/95 cases of limited intestinal infarction were observed, although 60% of patients had initial involvement of the superior mesenteric vein. Furthermore, recanalisation of the portal, splenic and superior mesenteric veins was obtained in 39%, 80%, and 73% of anticoagulated patients, respectively. Recanalisation of the portal vein did not occur in any of the patients beyond the sixth month of anticoagulation treatment. These findings independently validated retrospective single centre studies [[3], [40], [77], [78]]. Bleeding while on anticoagulation occurred in 9% of patients. Mortality rate was 2% and was not related to bleeding or PVT [3]. Among baseline factors, splenic vein obstruction and ascites [3] and delay in initiating anticoagulation [77] have been associated with the absence of recanalisation of the portal vein. These findings need further confirmation in other cohorts.

In most previous studies, anticoagulation therapy was mainly based on unfractionated heparin or LMWH or derivatives at high so-called therapeutic doses. In the most recent prospective European study, unfractionated heparin and LMWHs have been used in 25% and 65% of patients, respectively [3]. In most studies LMWH has been substituted for VKA targeting an INR between 2 and 3.

Heparin-induced thrombocytopenia (HIT) has been found to occur in up to 20% of PVT patients treated with unfractionated heparin, a much higher rate compared to HIT in patients without PVT [88]. The incidence is probably lower in patients treated with LMWH.


The experience of local thrombolysis, either venous or arterial, has been reported in no more than 100 patients, mainly as case reports. The transhepatic route or transjugular routes have been used. The reported recanalisation rates have been similar to those achieved with anticoagulation alone. However, 50% of treated patients developed major procedure-related bleeding, with a fatal outcome in some [[58], [89], [90]]. The transjugular approach for thrombolysis appears to be associated with reduced complications but the data remains limited to less than 30 treated patients [[91], [92]]. With surgical thrombectomy, recanalisation is achieved in only 30% of the patients. It is associated with a high recurrence rate, when performed >30 days from apparent onset [93]. Recently it has been shown that balloon angioplasty and/or stent placement without thrombolysis or thrombectomy may be a safe and effective treatment modality for post-operative main portal vein and superior mesenteric vein thrombosis [94]. As the long-term outcome of patients with chronic PVT is generally good (five-year survival rate above 70%) and mostly related to the associated conditions, the risk/benefit balance of such invasive procedures have to be considered [95].


When septic pylephlebitis is diagnosed, prolonged treatment with antibiotics adapted to isolated bacteria or to anaerobic digestive flora is necessary [96].


Recanalisation of the portal vein must be expected to occur up to 6 months whereas recanalisation of mesenteric and splenic veins steadily increase until 12 months follow-up [3]. Over half of the patients (55%) not achieving recanalisation will develop gastroesophageal varices during their follow-up, with a two-year actual probability of variceal bleeding of 12% and 16% for ascites [77]. Severe portal biliopathy, detected during imaging studies, developed in 30% of patients with acute PVT within 1 year [97].