Portal vein thrombosis may be seen in a variety of clinical contexts, and when acute can be a life-threatening condition. It is a major cause of non-cirrhotic presinusoidal portal hypertension. Portal vein thrombus may be either bland and/or malignant (i.e. tumour thrombus), and it is a critical finding in liver transplant candidates, as it precludes transplantation.
The demographics of patients with portal venous thrombosis will match those of the underlying condition.
Clinical presentation is often vague and non-specific. If extensive acute thrombosis is present, especially if the superior mesenteric venous system is also involved, then the presentation is likely to be with acute ischaemic bowel, mimicking superior mesenteric artery (SMA) occlusion.
Portal vein thrombosis, like thrombosis elsewhere, can occur due to disturbance of any one of the Virchow triad, and causes can be thought of in these terms 1,2:
- reduced flow / portal hypertension
- hypercoagulable state
- endothelial disturbance
Also, hepatocellular carcinoma (HCC) has a predilection for invading the portal vein, with tumour thrombus occluding the lumen 2.
Acutely only the thrombus may be evident, with associated findings related to the ischaemic bowel (especially if significant superior mesenteric venous (SMV) involvement is also present).
In chronic cases, cavernous transformation of the portal vein may be seen, with numerous periportal veins replacing the normal single channel of the portal vein.
Acute thrombosis may be difficult to detect with grey-scale imaging alone, as the thrombus may be hypoechoic. With time, it becomes more echogenic and easier to identify 3. Colour Doppler should be able to demonstrate absent flow in the portal vein and even to detect partial thrombosis, but attention to the Doppler gain and filters is necessary to avoid colour overwrite of partial thrombosis.
The SMV, intrahepatic branches of the portal vein, and hepatic veins should also be examined, to assess the extent of thrombosis (NB: 20% of patients with Budd-Chiari syndrome will also have portal vein thrombosis) 2.
Colour Doppler is also useful to help evaluate for tumour thrombus, which will show internal colour vascularity. Bland thrombus, in comparison, is avascular on colour Doppler.
Non-contrast scans are usually incapable of demonstrating the thrombus, except in some acute cases where the thrombus is hyperattenuating 2,4. In longstanding cases, low-density change in the liver may be evident, related to increased arterial supply, and representing fatty change.
The diagnosis can only reliably be made on portal venous phase contrast enhanced studies.
- complete or partial non-opacification of part of, or the whole, portal vein and its branches
- enhancement of the walls of the portal vein thought to represent either dilated vasa vasorum or a thin peripheral lumen remaining patent
When performing a hepatic CT protocol, the occlusion of a branch of the portal vein by a thrombus can manifest as transient hepatic attenuation differences (THAD) in the arterial/early portal phase, showing increased enhancement (i.e. perfusion) of the lobe or segment previously supplied by the vein due to hepatic arterial compensatory flow.
Importantly, the thrombus itself should not enhance. If enhancement is present, then this strongly suggests that the thrombus is not bland but rather represents tumour thrombus, most frequently from HCC 2.
Cavernous transformation appears as multiple small periportal vessels, which represent dilated collateral veins.
Associated findings of portal hypertension may, of course, be evident.
Although MRI is not as widely available, and can be difficult in unwell patients, it is the most sensitive modality for demonstrating portal venous thrombosis 2. 3D contrast enhanced MRA is the most sensitive sequence. However, the differentiation between bland and tumour thrombus usually requires integrating multiple sequences and taking into account chronicity of the thrombosis. Appearances include:
- acute thrombus will have high signal (see ageing blood on MRI)
- beware slow or turbulent flow artefacts
- acute thrombus may have high signal
- chronic thrombus may be low and appear as flow voids
- beware slow flow-related artefacts
- tumour thrombus is typically hyperintense
T1 + C (Gad)
- tumour thrombus enhancement may be detectable on postcontrast dynamic sequences
- transient hepatic intensity differences (THID) may be seen in the arterial/early portal phase if the thrombus occludes only a branch of the portal vein, as the hepatic artery takes on the affected lobe/segment's perfusion
- 3D contrast enhanced MRA (98% sensitive and 99% specific) 2
18F-FDG PET/CT has demonstrated a promising ability to differentiate between bland thrombus and malignant (tumour thrombus) portal vein thrombosis 5-11. The optimal imaging protocol appears to be with hybrid imaging part state-of-the-art CT including IV iodinated contrast media 10. Demonstrated features are:
- bland thrombus: no or slight FDG-avidity
- tumour thrombus: moderately to highly FDG-avid
- suggested optimal cutoff value: max SUV 2.3-3.6 8,11
- possible visualisation of further sites of thrombosis
Treatment and prognosis
Ideally, the presence of thrombosis is identified early and thrombolysis/thrombectomy performed to re-establish flow. This may avoid a possible long-term complication of portal hypertension, although portal vein thrombus often develops in a pre-existing setting of portal hypertension. Techniques that may be employed include 3:
- systemic anticoagulation
- endovascular infusion of thrombolytic agents: percutaneous transhepatic approach
- surgical thrombectomy
Even with treatment, half of affected patients go on to chronic occlusion 3. Surgical shunt formation may be useful for patient with portal hypertension.
The most important differential is to distinguish bland thrombus from tumour thrombus. Tumour thrombus is a contraindication to liver transplantation and is accorded a special LI-RADS status, LR5V (see: LI-RADS classification)
Porta hepatis masses (e.g. lymph nodes, cholangiocarcinomas) may compress the portal vein and may be confused for portal vein thrombus. They may, of course, co-exist with portal vein thrombosis.
Acute thrombosis, which is high attenuation on non-contrast CT, should be distinguished from other hyperattenuating masses in the region 4:
- 1. Janssen HL, Wijnhoud A, Haagsma EB et-al. Extrahepatic portal vein thrombosis: aetiology and determinants of survival. Gut. 2001;49 (5): 720-4. doi:10.1136/gut.49.5.720 - Free text at pubmed - Pubmed citation
- 2. Lee JK. Computed body tomography with MRI correlation. Lippincott Williams & Wilkins. (2006) ISBN:0781745268. Read it at Google Books - Find it at Amazon
- 3. Sheen CL, Lamparelli H, Milne A et-al. Clinical features, diagnosis and outcome of acute portal vein thrombosis. QJM. 2000;93 (8): 531-4. doi:10.1093/qjmed/93.8.531 - Pubmed citation
- 4. Mori H, Hayashi K, Uetani M et-al. High-attenuation recent thrombus of the portal vein: CT demonstration and clinical significance. Radiology. 1987;163 (2): 353-6. Radiology (abstract) - Pubmed citation
- 5. Parvey HR, Raval B, Sandler CM. Portal vein thrombosis: imaging findings. AJR Am J Roentgenol. 1994;162 (1): 77-81. AJR Am J Roentgenol (abstract) - Pubmed citation
- 6. Sun L, Guan YS, Pan WM et-al. Highly metabolic thrombus of the portal vein: 18F fluorodeoxyglucose positron emission tomography/computer tomography demonstration and clinical significance in hepatocellular carcinoma. World J. Gastroenterol. 2008;14 (8): 1212-7. Free text at pubmed - Pubmed citation
- 7. Sharma P, Kumar R, Singh H et-al. Imaging thrombus in cancer patients with FDG PET-CT. Jpn J Radiol. 2012;30 (2): 95-104. doi:10.1007/s11604-011-0016-9 - Pubmed citation
- 8. Lee EY, Khong PL. The value of 18F-FDG PET/contrast-enhanced CT in detection of tumor thrombus. Clin Nucl Med. 2013;38 (2): e60-5. doi:10.1097/RLU.0b013e318266d53e - Pubmed citation
- 9. Miceli M, Atoui R, Walker R et-al. Diagnosis of deep septic thrombophlebitis in cancer patients by fluorine-18 fluorodeoxyglucose positron emission tomography scanning: a preliminary report. J. Clin. Oncol. 2004;22 (10): 1949-56. doi:10.1200/JCO.2004.10.160 - Pubmed citation
- 10. Sopov V, Bernstine H, Stern D et-al. The metabolic spectrum of venous thrombotic disorders found on PET/CT. AJR Am J Roentgenol. 2009;193 (6): W530-9. doi:10.2214/AJR.09.2949 - Pubmed citation
- 11. Sharma P, Kumar R, Jeph S et-al. 18F-FDG PET-CT in the diagnosis of tumor thrombus: can it be differentiated from benign thrombus?. Nucl Med Commun. 2011;32 (9): 782-8. doi:10.1097/MNM.0b013e32834774c8 - Pubmed citation
- depositional disorders
- infection and inflammation
- liver abscess
- hepatic hydatid infection
- liver and intrahepatic bile duct tumours
- benign epithelial tumours
- hepatocellular hyperplasia
- hepatocellular adenoma
- hepatic/biliary cysts
- benign nonepithelial tumours
- primary malignant epithelial tumours
- primary malignant nonepithelial tumours
- haematopoietic and lymphoid tumours
- secondary tumours
- adrenal rest tumours
- hepatic carcinosarcoma
- hepatic fibroma
- hepatic Kaposi sarcoma
- hepatic lipoma
- hepatic mesenchymal hamartoma
- hepatic myxoma
- hepatic rhabdoid tumour
- hepatic solitary fibrous tumour
- hepatic teratoma
- hepatic yolk sac tumour
- inflammatory myofibroblastic tumour (inflammatory pseudotumor)
- nodular regenerative hyperplasia
- pancreatic rest tumours
- primary hepatic carcinoid
- benign epithelial tumours
- extrahepatic bile duct tumours
- liver and intrahepatic bile duct tumours
- portal venous gas
- portal hypertension
- portal vein thrombosis
- arterioportal shunts
- hereditary haemorrhagic telangiectasia (Osler-Weber-Rendu syndrome)
- Budd-Chiari syndrome
- passive hepatic congestion
- hepatic veno-occlusive disease
- hepatic infarction
- peliosis hepatis
- hepatic venous malformations (haemangiomas)
Ultrasound - liver
- ultrasound (introduction)
- liver ultrasound
- hepatic vasculature
- hepatic trauma on ultrasound
- liver transplant