Education + Advocacy = Change

 

Click a topic below for an index of articles:

New-Material

Home

Alternative-Treatments

Financial or Socio-Economic Issues

Forum

Health Insurance

Hepatitis

HIV/AIDS

Institutional Issues

International Reports

Legal Concerns

Math Models or Methods to Predict Trends

Medical Issues

Our Sponsors

Occupational Concerns

Our Board

Religion and infectious diseases

State Governments

Stigma or Discrimination Issues

 

IIf you would like to submit an article to this website, email us at info@heart-intl.net for a review of this paper
info@heart-intl.net

any words all words
Results per page:

“The only thing necessary for these diseases to the triumph is for good people and governments to do nothing.”


     

Hepatofugal Flow in the Portal Venous System: Pathophysiology, Imaging Findings, and Diagnostic Pitfalls1

Ronald H. Wachsberg, MD, Philip Bahramipour, MD, Constantine T. Sofocleous, MD and Allison Barone, MD

http://radiographics.rsnajnls.org/cgi/content/full/22/1/123

1 From the Department of Radiology, University of Medicine and Dentistry of New Jersey–New Jersey Medical School, Newark. Presented as an education exhibit at the 2000 RSNA scientific assembly. Received March 30, 2001; revision requested May 29 and received October 19; accepted October 19. Address correspondence to R.H.W., Department of Radiology, University Hospital, 150 Bergen St, Rm C-320, Newark, NJ 07103-2406 (e-mail: wachsbrh@umdnj.edu).


Abstract

 
Hepatofugal flow (ie, flow directed away from the liver) is abnormal in any segment of the portal venous system and is more common than previously believed. Hepatofugal flow can be demonstrated at angiography, Doppler ultrasonography (US), magnetic resonance imaging, and computed tomography (CT). The current understanding of hepatofugal flow recognizes the role of the hepatic artery and the complementary phenomena of arterioportal and portosystemic venovenous shunting. Detection of hepatofugal flow is clinically important for diagnosis of portal hypertension, for determination of portosystemic shunt patency and overall prognosis in patients with cirrhosis, as a potential pitfall at invasive arteriography performed to evaluate the patency of the portal vein, and as a contraindication to specialized imaging procedures (ie, transarterial hepatic chemoembolization and CT during arterial portography). Hepatofugal flow is generally diagnosed at Doppler US without much difficulty, but radiologists should beware of pitfalls that can impede correct determination of flow direction in the portal venous system.

© RSNA, 2002

Index Terms: Hypertension, portal, 957.711 • Liver, blood supply, 761.91 • Portal vein, flow dynamics, 957.453 • Shunts, arterioportal, 957.759 Shunts, portosystemic, 957.453


 

Introduction
 

 
Reversal of the normal direction of flow such that blood leaves an organ via the vessel that normally supplies it is unique to the portal venous system. Nearly two millennia ago, Galen posited that portal vein flow is hepatofugal (ie, directed away from the liver) unless food is present in the intestine, in which case portal vein flow is hepatopetal (ie, toward the liver) (1). It is now clear that blood flow in all branches of the normal portal venous system is always hepatopetal (Fig 1), which enables toxic substances absorbed in the intestine to be metabolized in the liver before entering the systemic circulation. The very existence of hepatofugal flow in patients with cirrhosis was the subject of debate until quite recently, and articles attesting to the phenomenon of hepatofugal flow were still being published only two decades ago (2–5). Owing principally to the advent of Doppler ultrasonography (US) over the past two decades, it is now recognized that hepatofugal flow is relatively common in patients with liver disease (5), and an appreciation of the complexity of this flow phenomenon has emerged in recent years (6).

 


Figure 1.   Normal portal venous circulation. Diagram shows hepatopetal flow (arrows) in all intrahepatic branches and extrahepatic tributaries of the portal venous system. In all of the diagrams in this article, a = artery and v = vein.

 
In this review, we discuss the imaging findings, differential diagnosis, clinical ramifications, and diagnostic pitfalls of hepatofugal flow in the portal venous system, with emphasis on Doppler US.


Imaging Findings

 
Depending on the involved segment(s) of the portal venous system, demonstration of hepatofugal flow is possible at angiography after injection of contrast material into the hepatic, superior mesenteric, or splenic artery (Fig 2); directly into the portal venous system (Fig 3); or into a hepatic vein (Fig 4). At Doppler US, hepatofugal flow appears as flow directed away from the liver in the portal vein, its intrahepatic branches, or its extrahepatic tributaries. If hepatofugal flow is present in the main portal vein or an intrahepatic branch, flow is noted in the direction opposite to flow in the adjacent hepatic artery (Fig 5). Depending on the anatomic orientation of the involved vessel relative to the transducer, a Doppler shift above or below the baseline can be produced by hepatofugal flow (Fig 6). A narrow portal vein and a prominent hepatic artery are common associated gray-scale US findings when flow is hepatofugal in the main portal vein (7,8). To-and-fro (bidirectional) blood flow, in which flow alternates between hepatopetal and hepatofugal during each cardiac cycle, has been observed to precede the development of frank hepatofugal flow in some patients with cirrhosis (Fig 7) and is the correlate of stagnant flow in the portal vein noted at arteriography (6,8).


Figures 2.   Hepatofugal flow seen with hepatic artery injection. Radiograph obtained during selective injection of contrast material into the right hepatic artery (arrow) shows opacification of the portal vein (PV) through an arterioportal fistula, which was believed to be a sequela of percutaneous liver biopsy.


Figures 3.   Hepatofugal flow seen with injection into the portal venous system in a patient with cirrhosis and recent hemorrhage from esophageal varices. Direct portogram obtained by injecting contrast material via a pigtail catheter (arrowhead), which was positioned in the proximal splenic vein (SPLV) with a transjugular approach, shows hepatofugal flow in the coronary vein (CV), leading to esophageal varices (*). PV = portal vein.


Figure 4.   Hepatofugal flow seen with hepatic vein injection in a patient with cirrhosis and portal hypertension. Retrograde portogram obtained by injecting carbon dioxide through a catheter, which was wedged in a peripheral hepatic vein (arrow) with a transjugular approach, shows retrograde opacification of the portal vein (PV), superior mesenteric vein (SMV), and coronary vein (CV) because of hepatofugal flow.



 

 

 

Figures 5.   Opposite flow directions in the portal vein and adjacent hepatic artery in a patient with cirrhosis and portal hypertension. Transverse color Doppler US image shows intrahepatic portal vein branches (white *) containing blue signal adjacent to hepatic artery branches (black *) containing red signal. Because blood flow is normally hepatopetal in both the portal vein and the hepatic artery, opposite color signals in adjacent branches of these two circulations indicate hepatofugal portal vein flow.

 

 


Figures 6.   Hepatofugal flow in a patient with cirrhosis and portal hypertension. Transverse color Doppler US image shows blue signal in both the right (RPV) and the left (LPV) intrahepatic branches of the portal vein. However, because of the anatomic configurations of these vessels relative to the ultrasound beam, flow is hepatopetal in the right portal vein but hepatofugal in the left portal vein.

 

 



 

 
Figure 7.   To-and-fro flow in a patient with cirrhosis and portal hypertension. Duplex US image of the left portal vein obtained during suspended respiration shows flow that alternates between hepatopetal (red arrowheads) and hepatofugal (blue arrowheads) with each heartbeat (white arrowheads). This transitional to-and-fro flow pattern is seen in some patients with cirrhosis before the development of frank hepatofugal flow.


 
Discrepancies are sometimes noted between findings at Doppler US and findings at arteriography with regard to flow direction in the portal vein (9). Although arteriography has long been regarded as the standard for determining the direction of portal vein flow, it has been shown that contrast material injected into the superior mesenteric artery provokes immediate vasodilatation, which causes a sudden increase in mesenteric venous return that can transiently convert portal vein flow from hepatofugal to hepatopetal during arteriography (10). In other patients, a sudden increase in mesenteric venous return following intraarterial contrast material injection can overload the portal vein and precipitate hepatofugal flow in the splenic vein during arteriography (11). Because invasive arteriography can disturb portal vein hemodynamics, Doppler US may be more accurate for determination of flow direction in the portal venous system (12).

Magnetic resonance (MR) angiography can be used to demonstrate hepatofugal flow in the portal venous system by using time-of-flight imaging with a traveling saturation pulse or phase-contrast techniques (13,14). However, mainly because of time constraints, angiography is not generally included in hepatic MR imaging protocols. Computed tomography (CT), although highly sensitive for detection of arterioportal fistulas, does not reliably demonstrate hepatofugal flow unless intense enhancement of a central portal vein branch is seen on postcontrast scans obtained during the early hepatic arterial phase (Fig 8) (15).

     



Figure 8.   Hepatofugal flow seen at CT in a patient with cirrhosis and diffuse hepatocellular carcinoma. Postcontrast CT scan obtained during the early arterial phase shows that the portal vein (*) is isoattenuating with the aorta. This finding indicates massive arterioportal shunting and hepatofugal flow in the portal vein, which was subsequently confirmed at Doppler US.

 

 Differential Diagnosis

 
Arterioportal Fistula
A direct communication between the hepatic artery and the portal vein can occur following trauma, in association with a focal liver lesion, or following rupture of a hepatic artery aneurysm into the portal vein or can be congenital (16). An arterioportal fistula is occasionally the sole cause of portal hypertension, but it is more often a sequela of invasive liver procedures (eg, biopsy, percutaneous transhepatic cholangiography) that are often performed in patients with hepatobiliary disease who may have preexisting portal hypertension. Because an arterioportal fistula is a potentially reversible condition that can exacerbate portal hypertension, one should routinely search for a fistula if hepatofugal flow is seen in an intrahepatic portal vein in a patient with cirrhosis (Fig 9) (17).

Figure 9a.   Arterioportal fistula in a patient with cirrhosis and recurrent hemorrhage from esophageal varices. (a) Transverse color Doppler US image shows rapid, turbulent flow in an arterioportal fistula (*), which has precipitated hepatofugal flow in the left portal vein (LPV). (b) Hepatic arteriogram shows the fistula (*), which is supplied by the hepatic artery (HA). Because of hepatofugal flow, contrast material opacifies the portal vein (PV) and superior mesenteric vein (SMV). After the fistula was ablated, portal vein flow became hepatopetal and the varices regressed.

 

Figure 9b.   Arterioportal fistula in a patient with cirrhosis and recurrent hemorrhage from esophageal varices. (a) Transverse color Doppler US image shows rapid, turbulent flow in an arterioportal fistula (*), which has precipitated hepatofugal flow in the left portal vein (LPV). (b) Hepatic arteriogram shows the fistula (*), which is supplied by the hepatic artery (HA). Because of hepatofugal flow, contrast material opacifies the portal vein (PV) and superior mesenteric vein (SMV). After the fistula was ablated, portal vein flow became hepatopetal and the varices regressed.

 
Depending on the size of an arterioportal fistula, the capacity of the draining portal vein can be exceeded by hepatic artery inflow, precipitating hepatofugal flow within the vein. In some cases, shunted hepatic artery blood joins the adjacent portal vein bloodstream without disturbing flow in more proximal portal vein branches (Fig 10). However, inflow from a large arterioportal fistula can overload the capacity of the intrahepatic portal venous system and precipitate hepatofugal flow in the main portal vein. In the latter instance, shunted hepatic artery blood is divided between two routes, with some traversing the sinusoids and exiting via hepatic veins whereas the remainder leaves the liver via the portal vein to reach the systemic circulation via portosystemic collateral vessels, so that the liver is perfused by arterial blood only (Fig 11) (5,7).


Figure 10a.   Small arterioportal shunt. (a) Diagram shows a hepatic lesion associated with a small arterioportal shunt. Shunted hepatic artery blood (striped red arrow) has precipitated hepatofugal flow in the portal vein branch that normally supplies the region where the lesion is situated. The shunted blood then joins normal hepatopetal blood flow in an adjacent portal vein branch (solid red arrow). Blue arrows = normal portal vein flow, green arrow = site of flow diversion. (b) Transverse color Doppler US image shows hepatofugal flow in the portal vein branch (arrows) that drains the lesion (m). Flow remains hepatopetal in more proximal portal veins (PV). The lesion had the typical appearance of a cavernous hemangioma at MR imaging and has remained stable for 6 years.



Figure 10b.   Small arterioportal shunt. (a) Diagram shows a hepatic lesion associated with a small arterioportal shunt. Shunted hepatic artery blood (striped red arrow) has precipitated hepatofugal flow in the portal vein branch that normally supplies the region where the lesion is situated. The shunted blood then joins normal hepatopetal blood flow in an adjacent portal vein branch (solid red arrow). Blue arrows = normal portal vein flow, green arrow = site of flow diversion. (b) Transverse color Doppler US image shows hepatofugal flow in the portal vein branch (arrows) that drains the lesion (m). Flow remains hepatopetal in more proximal portal veins (PV). The lesion had the typical appearance of a cavernous hemangioma at MR imaging and has remained stable for 6 years.



Figure 11a.   Large arterioportal shunt. (a) Diagram shows a hepatic tumor associated with a large arterioportal shunt that has precipitated hepatofugal flow (striped red arrows) in the left and main portal veins. Note that the hepatic artery provides the entire hepatic blood supply because splanchnic venous drainage does not enter the liver. Green arrows = site of flow diversion, solid red arrows = hepatopetal flow of shunted arterial blood. (b) Transverse color Doppler US image shows hepatofugal flow in the left and main portal vein (PV). The lesion was a hepatocellular carcinoma (not shown).



 

Figure 11b.   Large arterioportal shunt. (a) Diagram shows a hepatic tumor associated with a large arterioportal shunt that has precipitated hepatofugal flow (striped red arrows) in the left and main portal veins. Note that the hepatic artery provides the entire hepatic blood supply because splanchnic venous drainage does not enter the liver. Green arrows = site of flow diversion, solid red arrows = hepatopetal flow of shunted arterial blood. (b) Transverse color Doppler US image shows hepatofugal flow in the left and main portal vein (PV). The lesion was a hepatocellular carcinoma (not shown).
 

 
It was once believed that hepatofugal flow in a portal vein branch adjacent to a focal hepatic lesion was diagnostic for malignancy. However, arterioportal shunting is now recognized to occur in benign lesions as well (Fig 10) (18). Moreover, hepatofugal flow can occur in juxtalesional portal veins in the absence of intralesional shunting if the lesion compresses and obstructs regional hepatic venous outflow, thus leading to shunting of regional hepatic artery inflow into a regional portal vein, which is transformed into an outflow tract for the shunted blood (19,20). Similarly, it has been shown that regional portal veins are transformed into draining veins for hepatic artery inflow during occlusion of regional hepatic venous drainage (21). This phenomenon explains why hepatofugal flow is occasionally identified in portal vein branches adjacent to extraparenchymal lesions (eg, subcapsular hematoma) that compress hepatic venous drainage (14,15,22).

Portal Hypertension
The most common cause of hepatofugal flow in the portal venous system is portal hypertension, which in turn is usually caused by cirrhosis, less commonly by hepatic venous outflow obstruction or extrahepatic portal vein thrombosis. Portosystemic shunting in portal hypertension can be associated with hepatofugal flow in the main portal vein, intrahepatic branches only, or extrahepatic tributaries only, depending on the location of the shunt and the associated hemodynamic disturbances. The prevalence of hepatofugal flow in the portal venous system in studies of patients with cirrhosis evaluated with Doppler US varies between 3% and 23% (4,5,10,23,24). Discrepancies between series are mostly attributable to differences in the proportion of patients with advanced disease and to whether hepatofugal flow was evaluated in the main portal vein only versus in its major tributaries as well. In some patients with cirrhosis, hepatofugal flow is identified in isolated intrahepatic portal vein branches only, presumably because different regions of the liver are not equally affected by disease (8,25,26). Because the factors that influence flow direction in the portal vein are often in delicate equilibrium, a change in flow direction can be precipitated by even minor variation of one or more of these factors (Fig 12). Hepatofugal flow can change to hepatopetal flow after ingestion of a meal, presumably because of a postprandial increase in splanchnic venous flow and thus portal venous flow, although this phenomenon may be blunted in patients with cirrhosis (8,27). Hepatofugal flow can also revert to hepatopetal flow if the patient’s condition improves after medication is taken (5).


Figure 12a.   Iatrogenic change in flow direction in a patient with cirrhosis. Color Doppler US images of the right anterior portal vein (*) were obtained during a single breath hold. (a) Initial image shows hepatopetal flow. (b) Image obtained during application of gentle pressure by means of the transducer shows hepatofugal flow in the same vessel. This phenomenon was reproducible. The increase in transhepatic resistance caused by mild extrinsic pressure critically alters the delicate equilibrium between the determinants of flow direction in the portal vein.


Figure 12b.   Iatrogenic change in flow direction in a patient with cirrhosis. Color Doppler US images of the right anterior portal vein (*) were obtained during a single breath hold. (a) Initial image shows hepatopetal flow. (b) Image obtained during application of gentle pressure by means of the transducer shows hepatofugal flow in the same vessel. This phenomenon was reproducible. The increase in transhepatic resistance caused by mild extrinsic pressure critically alters the delicate equilibrium between the determinants of flow direction in the portal vein.

 
When hepatofugal flow occurs in extrahepatic portal vein tributaries only, splanchnic venous blood is shunted via portosystemic collateral vessels to the systemic circulation, whereas the intrahepatic circulation is not disturbed (Fig 13). However, when hepatofugal flow occurs in the main portal vein or intrahepatic branches, the pathophysiology is more complex. A comprehensive explanation of this phenomenon must account for the fact that high resistance to inflow in other organs (eg, a transplanted kidney with severe rejection) can precipitate thrombosis of the supply vessel but never causes flow reversal. Why does flow in the portal vein become hepatofugal in some patients with cirrhosis, instead of the portal vein simply becoming thrombosed in the face of high resistance to inflow?


Figure 13a.   Spontaneous splenorenal portosystemic shunt in a patient with cirrhosis. (a) Diagram shows shunting of a portion of superior mesenteric venous return to the left renal vein via the splenic vein. Note the hepatofugal flow (striped blue arrow) in the retropancreatic segment of the splenic vein. Green arrows = sites of flow diversion, solid blue arrows = normally directed venous flow. (b) Venous-phase superior mesenteric arteriogram shows the splenic vein (white *), superior mesenteric vein (black *), and portal vein (+). Opacification of the splenic vein is abnormal and indicates hepatofugal flow in this vessel. (c) Transverse color Doppler US image shows hepatofugal flow in the retropancreatic segment of the splenic vein, whereas the flow in the left renal vein is in the opposite direction. SMA = superior mesenteric artery.



Figure 13b.   Spontaneous splenorenal portosystemic shunt in a patient with cirrhosis. (a) Diagram shows shunting of a portion of superior mesenteric venous return to the left renal vein via the splenic vein. Note the hepatofugal flow (striped blue arrow) in the retropancreatic segment of the splenic vein. Green arrows = sites of flow diversion, solid blue arrows = normally directed venous flow. (b) Venous-phase superior mesenteric arteriogram shows the splenic vein (white *), superior mesenteric vein (black *), and portal vein (+). Opacification of the splenic vein is abnormal and indicates hepatofugal flow in this vessel. (c) Transverse color Doppler US image shows hepatofugal flow in the retropancreatic segment of the splenic vein, whereas the flow in the left renal vein is in the opposite direction. SMA = superior mesenteric artery.



 

Figure 13c.   Spontaneous splenorenal portosystemic shunt in a patient with cirrhosis. (a) Diagram shows shunting of a portion of superior mesenteric venous return to the left renal vein via the splenic vein. Note the hepatofugal flow (striped blue arrow) in the retropancreatic segment of the splenic vein. Green arrows = sites of flow diversion, solid blue arrows = normally directed venous flow. (b) Venous-phase superior mesenteric arteriogram shows the splenic vein (white *), superior mesenteric vein (black *), and portal vein (+). Opacification of the splenic vein is abnormal and indicates hepatofugal flow in this vessel. (c) Transverse color Doppler US image shows hepatofugal flow in the retropancreatic segment of the splenic vein, whereas the flow in the left renal vein is in the opposite direction. SMA = superior mesenteric artery.
 

 
The answer lies in the dual hepatic blood supply, wherein approximately 75% of hepatic blood flow normally arrives via the portal vein whereas the remainder is supplied by the hepatic artery (28). In cirrhosis, the principal locus of obstruction to blood flow is believed to be the hepatic venules and distal sinusoids (7,25). Because these are outflow vessels, one might anticipate that not only portal vein inflow but also hepatic artery inflow is impeded. Indeed, recent studies of patients with cirrhosis have documented abnormally high resistance to hepatic artery inflow that parallels the severity of portal hypertension (29–31). The term portal hypertension, although not a misnomer, is thus potentially misleading because it ignores the concomitant high resistance to inflow from the hepatic artery. In the setting of elevated resistance to portal vein inflow in cirrhosis, portosystemic collateral vessels spontaneously develop to decompress the portal vein (32), but there is no independent parallel mechanism by which the hepatic artery is decompressed. Rather, the hepatic artery "parasitizes" the portosystemic decompressive apparatus by gaining access to the portal venous system. This access is achieved by enlargement of normally minuscule communications between the hepatic arterial and portal venous circulations. Such communications have been identified in the hepatic sinusoids, the vasa vasorum of the portal vein, and peribiliary vascular plexuses (7,33,34). If transhepatic resistance to hepatic artery inflow is high relative to pressure gradients across intrahepatic arterioportal shunts and extrahepatic portosystemic collateral vessels, hepatic artery inflow will be shunted into (and can precipitate hepatofugal flow in) the portal vein and thereby gain access to the systemic circulation via portosystemic collateral vessels (Fig 14) (7,25). Thus, in patients with hepatofugal flow in the main portal vein or intrahepatic portal vein branches, the shunted blood originates in the hepatic artery. The critical role of the hepatic artery has been confirmed by Doppler US of patients with hepatofugal flow during intermittent occlusion of the hepatic artery (Fig 15) (35).



Figure 14.   Diffuse intrahepatic arterioportal shunting in a patient with cirrhosis. Diagram shows diffuse intrahepatic arterioportal shunting that drains via a portosystemic connection between the splenic and left renal veins. Note the hepatofugal flow (striped red arrows) in intrahepatic portal vein branches, the main portal vein, and the retropancreatic segment of the splenic vein. Superior mesenteric vein flow is also shunted via this pathway (striped blue arrow). Note that both hepatic artery flow and splanchnic venous flow are shunted to the systemic circulation via the splenorenal pathway, whereas only splanchnic venous blood is shunted via this pathway in the patient shown in Green arrows = sites of flow diversion, solid blue arrows = normally directed venous flow, solid red arrow = hepatic artery flow shunted via the splenorenal pathway.

 Figure 15a.   Change in portal vein flow direction during hepatic artery occlusion in a patient with cirrhosis that necessitated liver transplantation. (a) Intraoperative Doppler US image of the portal vein obtained during manual occlusion of the hepatic artery (before hepatectomy) shows hepatopetal flow. (b) Doppler US image obtained after discontinuation of hepatic artery occlusion shows that portal vein flow has become hepatofugal. This result demonstrates that hepatofugal flow in the portal vein is dependent on the hepatic artery. Note that the pulsations of the hepatic artery and portal vein are synchronous when portal vein flow is hepatofugal, a finding characteristic of arterioportal shunting.

 
Figure 15b.   Change in portal vein flow direction during hepatic artery occlusion in a patient with cirrhosis that necessitated liver transplantation. (a) Intraoperative Doppler US image of the portal vein obtained during manual occlusion of the hepatic artery (before hepatectomy) shows hepatopetal flow. (b) Doppler US image obtained after discontinuation of hepatic artery occlusion shows that portal vein flow has become hepatofugal. This result demonstrates that hepatofugal flow in the portal vein is dependent on the hepatic artery. Note that the pulsations of the hepatic artery and portal vein are synchronous when portal vein flow is hepatofugal, a finding characteristic of arterioportal shunting.
 

 
A transhepatic shunt is a unique type of portosystemic shunt that potentiates hepatopetal flow so that hepatofugal flow is virtually never seen in the main portal vein. A transhepatic shunt can be spontaneous (eg, a patent paraumbilical vein) or iatrogenic (eg, a transjugular intrahepatic portosystemic shunt [TIPS]). Such shunts are often associated with arterioportal shunting and hepatofugal flow in intrahepatic portal veins, so that the liver is paradoxically deprived of splanchnic venous blood despite brisk hepatopetal flow in the main portal vein. In one large series, hepatofugal flow in some or all intrahepatic portal vein branches was observed in 29% of patients with a patent paraumbilical vein, and such patients had significantly larger esophageal varices than patients with exclusively hepatopetal flow in intrahepatic portal veins (36). In patients with a patent paraumbilical vein, hepatofugal flow can be seen in only left portal vein branches (ie, adjacent to the paraumbilical vein) (Fig 16) or throughout the entire liver (Fig 17) (36). Following TIPS creation, intrahepatic hemodynamics mirror the physiology seen in patients with a patent paraumbilical vein, with the important difference that hepatofugal flow in intrahepatic portal veins is a desirable finding that is seen in the great majority of patients with a properly functioning TIPS (Fig 18) (37).

Figure 16.   Patent paraumbilical vein in a patient with cirrhosis. Transverse color Doppler US image shows hepatofugal flow in the portal vein branch to segment 4 (arrowheads), which indicates localized arterioportal shunting. Flow in the main portal vein and right portal vein branches was hepatopetal.
 

 


Figure 17a.   Prominent paraumbilical vein in a patient with cirrhosis. (a) Diagram shows a large paraumbilical vein associated with hepatopetal flow in the main portal vein but hepatofugal flow (striped red arrows) in intrahepatic portal vein branches. Both splanchnic venous blood (blue arrows) and hepatic artery blood are shunted to the systemic venous circulation via the paraumbilical vein. Despite hepatopetal flow in the portal vein, the hepatic parenchyma is not perfused by splanchnic venous blood because portal venous inflow is completely shunted to the paraumbilical vein. Solid red arrow = hepatic artery blood shunted via the paraumbilical vein. (b) Transverse color Doppler US image shows hepatofugal flow in the right portal vein (PV) and hepatopetal flow in the main portal vein. (c) Postcontrast CT scan obtained during the arterial phase shows that the paraumbilical vein (black *) is enhanced, whereas the portal vein (white *) has not yet enhanced. This finding indicates that blood arriving via the hepatic artery has traversed arterioportal shunts to enter intrahepatic portal veins and thereby gain access to the systemic venous system via the paraumbilical vein.

 

Figure 17b.   Prominent paraumbilical vein in a patient with cirrhosis. (a) Diagram shows a large paraumbilical vein associated with hepatopetal flow in the main portal vein but hepatofugal flow (striped red arrows) in intrahepatic portal vein branches. Both splanchnic venous blood (blue arrows) and hepatic artery blood are shunted to the systemic venous circulation via the paraumbilical vein. Despite hepatopetal flow in the portal vein, the hepatic parenchyma is not perfused by splanchnic venous blood because portal venous inflow is completely shunted to the paraumbilical vein. Solid red arrow = hepatic artery blood shunted via the paraumbilical vein. (b) Transverse color Doppler US image shows hepatofugal flow in the right portal vein (PV) and hepatopetal flow in the main portal vein. (c) Postcontrast CT scan obtained during the arterial phase shows that the paraumbilical vein (black *) is enhanced, whereas the portal vein (white *) has not yet enhanced. This finding indicates that blood arriving via the hepatic artery has traversed arterioportal shunts to enter intrahepatic portal veins and thereby gain access to the systemic venous system via the paraumbilical vein.

 
Figure 17c.   Prominent paraumbilical vein in a patient with cirrhosis. (a) Diagram shows a large paraumbilical vein associated with hepatopetal flow in the main portal vein but hepatofugal flow (striped red arrows) in intrahepatic portal vein branches. Both splanchnic venous blood (blue arrows) and hepatic artery blood are shunted to the systemic venous circulation via the paraumbilical vein. Despite hepatopetal flow in the portal vein, the hepatic parenchyma is not perfused by splanchnic venous blood because portal venous inflow is completely shunted to the paraumbilical vein. Solid red arrow = hepatic artery blood shunted via the paraumbilical vein. (b) Transverse color Doppler US image shows hepatofugal flow in the right portal vein (PV) and hepatopetal flow in the main portal vein. (c) Postcontrast CT scan obtained during the arterial phase shows that the paraumbilical vein (black *) is enhanced, whereas the portal vein (white *) has not yet enhanced. This finding indicates that blood arriving via the hepatic artery has traversed arterioportal shunts to enter intrahepatic portal veins and thereby gain access to the systemic venous system via the paraumbilical vein.
 

 

 


Figure 18a.   TIPS in a patient with cirrhosis. (a) Diagram shows hepatofugal flow in intrahepatic portal veins (striped red arrows) and hepatopetal flow in the main portal vein. Note the similarity to the hemodynamics seen with a large paraumbilical vein (Fig 17a). Blue arrows = splanchnic venous blood, solid red arrow = hepatic artery blood shunted via a TIPS. (b) Postcontrast CT scan obtained during the arterial phase shows early, intense enhancement of the left portal vein (PV), which indicates hepatofugal flow in this vessel. * = the TIPS. (c) Oblique color Doppler US image shows hepatopetal flow in the main portal vein (PV) and hepatofugal flow in the right anterior portal vein.
 

 

 

 


Figure 18b.   TIPS in a patient with cirrhosis. (a) Diagram shows hepatofugal flow in intrahepatic portal veins (striped red arrows) and hepatopetal flow in the main portal vein. Note the similarity to the hemodynamics seen with a large paraumbilical vein (Fig 17a). Blue arrows = splanchnic venous blood, solid red arrow = hepatic artery blood shunted via a TIPS. (b) Postcontrast CT scan obtained during the arterial phase shows early, intense enhancement of the left portal vein (PV), which indicates hepatofugal flow in this vessel. * = the TIPS. (c) Oblique color Doppler US image shows hepatopetal flow in the main portal vein (PV) and hepatofugal flow in the right anterior portal vein.
 

 

 


Figure 18c.   TIPS in a patient with cirrhosis. (a) Diagram shows hepatofugal flow in intrahepatic portal veins (striped red arrows) and hepatopetal flow in the main portal vein. Note the similarity to the hemodynamics seen with a large paraumbilical vein (Fig 17a). Blue arrows = splanchnic venous blood, solid red arrow = hepatic artery blood shunted via a TIPS. (b) Postcontrast CT scan obtained during the arterial phase shows early, intense enhancement of the left portal vein (PV), which indicates hepatofugal flow in this vessel. * = the TIPS. (c) Oblique color Doppler US image shows hepatopetal flow in the main portal vein (PV) and hepatofugal flow in the right anterior portal vein.
 

 

     



Clinical Ramifications

 
Diagnosis of Portal Hypertension
In a patient with known or suspected cirrhosis, it is important to determine whether portal hypertension is present. With few exceptions, hepatofugal flow in the main portal vein or an extrahepatic portal vein tributary is a specific sign of portal hypertension. One exception is hepatofugal flow in a liver transplant recipient with a large, persistent portosystemic collateral vessel that can divert a substantial amount of splanchnic venous blood; this diversion can interfere with graft function and portal vein patency but does not indicate recurrent portal hypertension (Fig 19) (38). Another rare exception is hepatofugal flow caused by a congenital portosystemic collateral vessel (39). Hepatofugal flow in one or more solely intrahepatic portal veins can occur in patients with a focal arterioportal shunt and is therefore not specific for portal hypertension (19).

 


Figure 19a.   Graft dysfunction in a recent recipient of an orthotopic liver transplant. (a) Transverse duplex Doppler US image shows hepatofugal flow in the splenic vein. Flow was hepatopetal in the portal vein. Note the dilated left renal vein (*), which is consistent with splenorenal portosystemic shunting. A = aorta, P = pancreas. (b) Axial T1-weighted spin-echo MR image shows a prominent splenorenal collateral vessel (arrow) and a dilated left renal vein (*). Invasive studies showed a normal portohepatic gradient, which indicated that the hepatofugal flow was not caused by portal hypertension. Graft function normalized after the collateral vessel was surgically ligated.

 

Figure 19b.   Graft dysfunction in a recent recipient of an orthotopic liver transplant. (a) Transverse duplex Doppler US image shows hepatofugal flow in the splenic vein. Flow was hepatopetal in the portal vein. Note the dilated left renal vein (*), which is consistent with splenorenal portosystemic shunting. A = aorta, P = pancreas. (b) Axial T1-weighted spin-echo MR image shows a prominent splenorenal collateral vessel (arrow) and a dilated left renal vein (*). Invasive studies showed a normal portohepatic gradient, which indicated that the hepatofugal flow was not caused by portal hypertension. Graft function normalized after the collateral vessel was surgically ligated.

 

 
Evaluation of Portosystemic Shunt Function
Surgical Shunts.— Because overlying structures can interfere with direct imaging of a portosystemic shunt at US, hepatofugal flow in the portal vein and extrahepatic tributaries leading to the shunt is an important indirect sign of shunt patency (Fig 20) (40). This finding is not usually seen in patients with a reduced-diameter or selective shunt (41,42).

Figures 20.   Surgical mesocaval shunt in a patient with cirrhosis. Longitudinal color Doppler US image shows hepatofugal flow in the portal vein (PV) and superior mesenteric vein (SMV), which is strong ancillary evidence of shunt patency; the shunt itself was obscured by bowel gas. The arrows indicate the direction of flow. HA = hepatic artery.

 
TIPS.— Flow in intrahepatic portal vein branches is hepatofugal in most patients following TIPS creation, as noted earlier (37). Indeed, the presence of hepatopetal flow in intrahepatic portal vein branches is highly predictive of TIPS malfunction (Fig 21), particularly if hepatofugal flow had been previously documented in the affected vessel (37,43).


 

Figures 21.   TIPS in a patient with portal hypertension. Color Doppler US image shows hepatopetal flow in the left portal vein (PV), even though flow velocities within the TIPS and the portal vein were within the expected range. As a result of this abnormal finding, venography was performed, which demonstrated a substantial portosystemic gradient caused by pseudointimal hyperplasia.

 
Prognostication
Patients with cirrhosis and hepatofugal flow in the main portal vein or intrahepatic portal vein branches have a poorer prognosis and are more likely to have severe manifestations of liver disease than patients with hepatopetal portal vein flow (24,26). There is evidence that patients with cirrhosis are at relatively low risk for hemorrhage from esophageal varices if flow in the coronary vein is hepatopetal, even if other complications of end-stage disease have developed (14,44,45). In patients with hepatocellular carcinoma, life expectancy is substantially shorter if portal vein flow is hepatofugal than if portal vein flow is hepatopetal, presumably because extensive arterioportal shunting is more likely in aggressive tumors (45).

Cause of False-Positive Angiographic Diagnosis of Portal Vein Thrombosis
In patients with hepatofugal flow, the portal vein does not opacify after contrast material injection into the superior mesenteric or splenic artery because splanchnic venous return does not enter the portal vein. Portal vein thrombosis can be mistakenly diagnosed if this fact is not borne in mind (Fig 22).

 

 

Figure 22.   False-positive arteriographic diagnosis of portal vein occlusion in a liver transplant recipient with graft dysfunction. Venous-phase superior mesenteric arteriogram shows that the inferior vena cava (IVC), the superior mesenteric vein (SMV), the splenic vein (SPLV), and varices (*) are opacified. Note the cutoff (arrow) where the portal vein should be seen, a finding that suggests portal vein thrombosis. Doppler US performed immediately afterward revealed hepatofugal portal vein flow, which indicated a steal syndrome caused by large, persistent splenorenal collateral vessels
 

 
Contraindication to Specialized Imaging Procedures
Transarterial chemoembolization of hepatocellular carcinoma is contraindicated in patients with hepatofugal flow in the portal vein, because medications administered via the hepatic artery do not enter the tumor but are instead shunted to the systemic circulation (Fig 23) (46). CT during arterial portography is a technique that maximizes the conspicuity of liver tumors by causing the portal vein to enhance earlier than the hepatic artery. To achieve this, contrast material must be administered through a catheter positioned in the splenic or superior mesenteric artery. The parenchymal liver enhancement is inadequate to justify this procedure in patients with hepatofugal flow, in whom a substantial proportion of enhanced venous return is shunted away from the liver to the systemic circulation (47).


 

Figure 23.   Arterioportal shunt in a patient with hepatocellular carcinoma in the left lobe who was referred for transarterial chemoembolization. Hepatic arteriogram shows opacification of the right and main portal veins (PV) via a large intratumoral fistula (*). Extensive arterioportal shunting contraindicated the planned procedure because the chemotherapeutic agent would have been shunted to the systemic circulation rather than penetrating the tumor.

 
Diagnostic Pitfalls
 

 
A variety of pitfalls can cause underdiagnosis or overdiagnosis of hepatofugal flow at Doppler US (48,49). It is important to carefully ascertain the identity of the insonated vessel of interest before concluding that hepatofugal flow is present (Figs 24, 25). One cannot assume that the largest hepatic blood vessels are portal veins, because the hepatic artery and its branches are often quite prominent in patients with liver disease (Fig 26) (8). One should also be aware that hepatofugal flow can develop in intrahepatic arterial branches if arterial collateral vessels have developed after hepatic artery thrombosis in a liver transplant recipient (Fig 27) (50). Aliasing can be misinterpreted as hepatofugal flow if the pulse repetition frequency setting is inappropriately low (Figs 28, 29). Helical portal vein flow, a recognized mimic of hepatofugal flow, is particularly common in liver transplant recipients when the donor portal vein is substantially wider than the recipient portal vein (51). Because the telltale corkscrew appearance may not be readily obvious at color Doppler US, helical flow should be carefully excluded if findings suggest hepatofugal flow in a liver transplant recipient (Figs 30, 31).


Figure 24a.   False-positive diagnosis of hepatofugal flow in a patient with cirrhosis and portal hypertension. (a) Color Doppler US image shows hepatopetal flow in the right portal vein (+) and hepatofugal flow in what appears to be the main portal vein (*). Note the similarity to the findings seen with a large paraumbilical vein (cf Fig 4). (b) Color Doppler US image obtained after repositioning of the transducer shows hepatopetal flow in the main portal vein (PV). The apparent hepatofugal flow in a was in fact hepatopetal flow in the left portal vein, which was mistakenly identified as the main portal vein.

 



 

Figure 24b.   False-positive diagnosis of hepatofugal flow in a patient with cirrhosis and portal hypertension. (a) Color Doppler US image shows hepatopetal flow in the right portal vein (+) and hepatofugal flow in what appears to be the main portal vein (*). Note the similarity to the findings seen with a large paraumbilical vein (cf Fig 4). (b) Color Doppler US image obtained after repositioning of the transducer shows hepatopetal flow in the main portal vein (PV). The apparent hepatofugal flow in a was in fact hepatopetal flow in the left portal vein, which was mistakenly identified as the main portal vein.
 

 

 

 

 

Figure 25.   False-positive diagnosis of hepatofugal flow in a healthy subject. Color Doppler US image shows two pairs of parallel vessels in the left hepatic lobe. For each vascular pair, flow in one vessel is toward the transducer (black *) and flow in the adjacent vessel is in the opposite direction (white *). (Note the similarity to the findings in the patient with true hepatofugal flow shown in Figure 5.) What is actually seen in this image is normal flow in portal and hepatic vein branches, which are often parallel to each other in the periphery of the liver.
 

 


Figure 26a.   Prominent hepatic artery branches in a patient with cirrhosis and portal hypertension. (a) Oblique color Doppler US image shows hepatopetal flow in prominent intrahepatic vessels (arrowheads), which were initially thought to be portal vein branches. (b) Duplex US image shows that the prominent intrahepatic vessels are hepatic artery branches. Hepatofugal flow in a small portal vein (arrow) was nearly overlooked. Although portal vein branches are usually much wider than adjacent hepatic artery branches, the latter can be quite prominent relative to portal vein branches in patients with cirrhosis.
 

 

Figure 26b.   Prominent hepatic artery branches in a patient with cirrhosis and portal hypertension. (a) Oblique color Doppler US image shows hepatopetal flow in prominent intrahepatic vessels (arrowheads), which were initially thought to be portal vein branches. (b) Duplex US image shows that the prominent intrahepatic vessels are hepatic artery branches. Hepatofugal flow in a small portal vein (arrow) was nearly overlooked. Although portal vein branches are usually much wider than adjacent hepatic artery branches, the latter can be quite prominent relative to portal vein branches in patients with cirrhosis.

 

 

 

 


Figure 27.   Hepatofugal flow in the hepatic artery in a liver transplant recipient with elevated liver enzyme levels. Color Doppler US image shows flow away from the transducer in one vessel (white *) and flow toward the transducer in the adjacent vessel (black *). Doppler spectrum shows venous flow above the baseline and arterial flow below it; these findings indicate that portal vein flow is hepatopetal whereas hepatic artery flow is hepatofugal. Arterial collateral vessels formed after hepatic artery thrombosis; thus, arterial blood enters the liver via anterior transcapsular vessels and flows toward the porta hepatis. These findings are similar to those in the patient with hepatofugal portal vein flow in Figure 5.
 

 

Figure 28a.   Aliasing artifact in a patient with cirrhosis. The patient was referred for surveillance of a TIPS (*). (a) Color Doppler US image shows blue signal in the portal vein (PV), which is consistent with hepatofugal flow. However, flow in the boundary layer (adjacent to the wall of the portal vein) is shown as red signal, which implausibly indicates flow in a direction opposite to that of flow in the central lumen. (b) Color Doppler US image obtained after the pulse repetition frequency (scale) was increased shows only red signal in the lumen of the portal vein (PV). This finding indicates hepatopetal flow and confirms that aliasing artifact accounts for the blue color signal in a.

 

 


Figure 28b.   Aliasing artifact in a patient with cirrhosis. The patient was referred for surveillance of a TIPS (*). (a) Color Doppler US image shows blue signal in the portal vein (PV), which is consistent with hepatofugal flow. However, flow in the boundary layer (adjacent to the wall of the portal vein) is shown as red signal, which implausibly indicates flow in a direction opposite to that of flow in the central lumen. (b) Color Doppler US image obtained after the pulse repetition frequency (scale) was increased shows only red signal in the lumen of the portal vein (PV). This finding indicates hepatopetal flow and confirms that aliasing artifact accounts for the blue color signal in a.
 

 


Figure 29a.   Aliasing artifact caused by a stenotic web at the portal vein anastomosis in a liver transplant recipient. (a) Color Doppler US image appears to show an abrupt transition from hepatopetal flow to hepatofugal flow (*) in the portal vein. Note that red signal is not seen in the boundary layer of the portal vein (unlike in the patient shown in Fig 28). Also, note the similarity to the findings in the patient with a patent paraumbilical vein shown in Figure 17b. (b) Gray-scale US image shows that an intraluminal web (arrows) is the cause of the poststenotic color aliasing.
 

 

 

Figure 29b.   Aliasing artifact caused by a stenotic web at the portal vein anastomosis in a liver transplant recipient. (a) Color Doppler US image appears to show an abrupt transition from hepatopetal flow to hepatofugal flow (*) in the portal vein. Note that red signal is not seen in the boundary layer of the portal vein (unlike in the patient shown in Fig 28). Also, note the similarity to the findings in the patient with a patent paraumbilical vein shown in Figure 17b. (b) Gray-scale US image shows that an intraluminal web (arrows) is the cause of the poststenotic color aliasing.

 

 

 


Figures 30.   Helical flow in an adult recipient of a whole-liver transplant. Color Doppler US image shows alternating bands of blue and red signal (arrowheads), which indicate helical flow in the portal vein, a common finding after transplantation. This characteristic appearance is caused by a continuously changing Doppler angle when blood flows in a helical trajectory (46). Exclusively hepatopetal flow (*) within the liver confirms that the net flow is indeed hepatopetal.
 

 
Figures 31.   Helical flow in a pediatric recipient of a reduced-liver transplant. Color Doppler US image shows predominantly blue signal (white *) in the portal vein, which suggests hepatofugal flow. Minimal red signal at the periphery of the portal vein (black *) might be overlooked. However, flow in intrahepatic portal vein branches was exclusively hepatopetal; this finding indicated that helical flow is responsible for the color Doppler US findings in the hilar portal vein. This deceptive appearance is the reason why helical flow should be considered when Doppler US shows apparent hepatofugal flow, even if the classic corkscrew appearance is not seen. Note the prominent hepatic artery (HA) on either side of the portal vein.
 

 
Conclusions
 

 
Hepatofugal flow in the portal venous system is a complex flow abnormality seen in patients with focal and diffuse hepatic disease. The presence of hepatofugal flow has several important clinical implications. Hepatofugal flow is generally diagnosed at Doppler US without much difficulty, but imagers should beware of pitfalls that can impede correct determination of flow direction in the portal venous system.


Footnotes

 
Abbreviation: TIPS = transjugular intrahepatic portosystemic shunt


References

 

  1. Granger DN, Shepherd AP. The intestinal circulation: a historical perspective. In: Shepherd AP, Granger DN, eds. Physiology of the intestinal circulation. New York, NY: Raven, 1984; 1-8.
  1. Moreno AH, Burchell AR, Reddy RV, Steen JA, Panke WF, Nealson TF, Jr. Spontaneous reversal of portal blood flow: the case for and against its occurrence in patients with cirrhosis of the liver. Ann Surg 1975; 181:346-358.
  1. Takayasu K, Takashi M, Musha H, et al. Spontaneous reversal of portal blood flow demonstrated by percutaneous transhepatic catheterization. Gastroenterology 1982; 82:753-757.
  1. Burcharth F, Aagaard J. Total hepatofugal flow in cirrhosis demonstrated by transhepatic portography. ROFO Fortschr Geb Rontgenstr Nuklearmed 1988; 148:47-49.
  1. Kawasaki T, Moriyasu F, Nishida O, et al. Analysis of hepatofugal flow in portal venous system using ultrasonic Doppler duplex system. Am J Gastroenterol 1989; 84:937-941.
  1. Rector WG, Jr, Hoefs JC, Hossack KF, Everson GT. Hepatofugal portal flow in cirrhosis: observations on hepatic hemodynamics and the nature of the arterioportal communications. Hepatology 1988; 8:16-20.
  1. Bookstein JJ, Cho KJ, Becker GB, Dail D. Arterioportal communications: observations and hypotheses concerning transsinusoidal and transvasal types. Radiology 1982; 142:581-590.
  1. Ralls PW. Color Doppler sonography of the hepatic artery and portal venous system. AJR Am J Roentgenol 1990; 155:517-525.
  1. Nelson RC, Lovett KE, Chezmar JL, et al. Comparison of pulsed Doppler sonography and angiography in patients with portal hypertension. AJR Am J Roentgenol 1987; 149:77-81.
  1. Gaiani S, Bolondi L, Li Bassi S, Zironi G, Siringo S, Barbara L. Prevalence of spontaneous hepatofugal portal flow in liver cirrhosis: clinical and endoscopic correlation in 228 patients. Gastroenterology 1991; 100:160-167.
  1. Ohnishi K, Saito M, Sato S, et al. Direction of splenic venous flow assessed by pulsed Doppler flowmetry in patients with a large splenorenal shunt: relation to spontaneous hepatic encephalopathy. Gastroenterology 1985; 89:180-185.
  1. Gaiani S, Bolondi L, Barbara L. Prevalence of spontaneous hepatofugal portal flow in liver cirrhosis (letter). Gastroenterology 1991; 101:1142.
  1. Burkart DJ, Johnson CD, Morton MJ, Ehman RL. Phase-contrast cine MR angiography in chronic liver disease. Radiology 1993; 187:407-412.
  1. Imazu H, Matsui T, Noguchi R, et al. Magnetic resonance angiography for monitoring prophylactic endoscopic treatment of high risk esophageal varices. Endoscopy 2000; 32:766-772.
  1. Park CM, Cha SH, Kim DH, et al. Hepatic arterioportal shunts not directly related to hepatocellular carcinoma: findings on CT during hepatic arteriography, CT arterial portography and dual phase spiral CT. Clin Radiol 2000; 55:465-470.
  1. Vauthey JN, Tomczak RJ, Helmberger T, et al. The arterioportal fistula syndrome: clinicopathologic features, diagnosis, and therapy. Gastroenterology 1997; 113:1390-1401.
  1. Bolognesi M, Sacerdoti D, Bombonato G, Chiesura-Corona M, Merkel C, Gatta A. Arterioportal fistulas in patient with liver cirrhosis: usefulness of color Doppler US for screening. Radiology 2000; 216:738-743.
  1. Naganuma H, Ishida H, Konno K, et al. Hepatic hemangioma with arterioportal shunts. Abdom Imaging 1999; 24:42-46.
  1. Miller MA, Balfe DM, Middleton WD. Peripheral portal venous blood flow alterations induced by hepatic masses: evaluation with color and pulsed Doppler sonography. J Ultrasound Med 1996; 15:707-713.
  1. Lin ZY, Wang JH, Wang LY, et al. Changes in intrahepatic portal hemodynamic in early stage hepatic abscesses. J Ultrasound Med 1996; 15:595-598.
  1. Murata S, Itai Y, Asato M, et al. Effect of temporary occlusion of the hepatic vein on dual blood in the liver: evaluation with spiral CT. Radiology 1995; 197:351-356.
  1. Pearl LB, Trunkey DD. Compartment syndrome of the liver. J Trauma 1999; 47:796-798.
  1. Taourel P, Blanc P, Dauzat M, et al. Doppler study of mesenteric, hepatic, and portal circulation in alcoholic cirrhosis: relationship between quantitative Doppler measurements and the severity of portal hypertension and hepatic failure. Hepatology 1998; 28:932-936.
  1. von Herbay A, Frieling T, Haussinger D. Color Doppler sonographic evaluation of spontaneous portosystemic shunts and inversion of portal venous flow in patients with cirrhosis. J Clin Ultrasound 2000; 28:332-339.
  1. Hoevels J, Lunderquist A, Tylen U. Spontaneous intermittent reversal of blood flow in intrahepatic portal vein branches in cirrhosis of the liver. Cardiovasc Radiol 1979; 2:267-273.
  1. Tarantino L, Giorgio A, de Stefano G, et al. Reverse flow in intrahepatic portal vessels and liver function impairment in cirrhosis. Eur J Ultrasound 1997; 6:171-177.
  1. de Vries PJ, de Hooge P, Hoekstra JB, van Hattum J. Blunted postprandial reaction of portal venous flow in chronic liver disease, assessed with duplex Doppler: significance for prognosis. J Hepatol 1994; 21:966-973.
  1. Lautt WW, Greenway CV. Conceptual review of the hepatic vascular bed. Hepatology 1987; 7:952-963.
  1. Sacerdoti D, Merkel C, Bolognesi M, Amodio P, Angeli P, Gatta A. Hepatic arterial resistance in cirrhosis with and without portal vein thrombosis: relationships with portal hemodynamics. Gastroenterology 1995; 108:1152-1158.
  1. Joynt LK, Platt JF, Rubin JM, Ellis JH, Bude RO. Hepatic artery resistance before and after standard meal in subjects with diseased and healthy livers. Radiology 1995; 196:489-492.
  1. Piscaglia F, Gaiani S, Zironi G, et al. Intra- and extrahepatic arterial resistances in chronic hepatitis and liver cirrhosis. Ultrasound Med Biol 1997; 23:675-682.
  1. Cho KC, Patel YD, Wachsberg RH, Seeff J. Varices in portal hypertension: evaluation with CT. RadioGraphics 1995; 15:609-622.
  1. Kan Z, Ivancev K, Lunderquist A. Peribiliary plexa: important pathways for shunting of iodized oil and silicone rubber solution from the hepatic artery to the portal vein—an experimental study in rats. Invest Radiol 1994; 7:671-676.
  1. Quiroga S, Sebastia C, Pallisa E, Castella E, Perez-Lafuente M, Alvarez-Castells A. Improved diagnosis of hepatic perfusion disorders: value of hepatic arterial phase imaging during helical CT. RadioGraphics 2001; 21:65-81.
  1. Wachsberg RH, Needleman L, Wilson DJ. Portal vein pulsatility in normal and cirrhotic adults without cardiac disease. J Clin Ultrasound 1995; 23:3-15.
  1. Morin C, Lafortune M, Pomier G, Robin M, Breton G. Patent paraumbilical vein: anatomic and hemodynamic variants and their clinical importance. Radiology 1992; 185:253-256.
  1. Kanterman RY, Darcy MD, Middleton WD, Sterling KM, Teefey SA, Pilgram TK. Doppler sonography findings associated with transjugular intrahepatic portosystemic shunt malfunction. AJR Am J Roentgenol 1997; 168:467-472.
  1. Durham JD, LaBerge JM, Altman S, et al. Portal vein thrombolysis and closure of competitive shunts following liver transplantation. J Vasc Interv Radiol 1994; 5:611-615.
  1. Nishimoto Y, Hoshino H, Sato S, et al. Extrahepatic portosystemic venous shunt without portal hypertension. Intern Med 1997; 36:886-889.
  1. Lafortune M, Patriquin H, Pomier G, et al. Hemodynamic changes in portal circulation after portosystemic shunts: use of Doppler sonography in 43 patients. AJR Am J Roentgenol 1987; 149:701-706.
  1. Lacy AM, Navasa M, Gilbert R, et al. Long-term effects of distal splenorenal shunt on hepatic hemodynamics and liver function in patients with cirrhosis: importance of reversal of portal blood flow. Hepatology 1992; 15:616-622.
  1. Rosemurgy AS, McAllister EW, Goode SE. Direction or reversal of preshunt portal blood flow as determinants of outcome up to 1 year after small-diameter prosthetic H-graft portacaval shunt. J Surg Res 1995; 58:432-434.
  1. Feldstein VA, Patel MD, LaBerge JM. Transjugular intrahepatic portosystemic shunts: accuracy of Doppler US in determination of patency and detection of stenoses. Radiology 1996; 201:141-147.
  1. Wachsberg RH, Simmons MZ. Coronary vein diameter and flow direction in patients with portal hypertension: evaluation with duplex sonography and correlation with variceal bleeding. AJR Am J Roentgenol 1994; 162:637-641.
  1. Komatsuda T, Ishida H, Konno K, et al. Color Doppler findings of gastrointestinal varices. Abdom Imaging 1998; 23:45-50.
  1. Ngan H, Peh WC. Arteriovenous shunting in hepatocellular carcinoma: its prevalence and clinical significance. Clin Radiol 1997; 52:36-40.
  1. Oliver JH, Baron RL, Dodd GD, Peterson MS, Carr BI. Does advanced cirrhosis with portosystemic shunting affect the value of CT arterial portography in the evaluation of the liver?. AJR Am J Roentgenol 1995; 164:333-337.
  1. Pozniak MA, Zagzebski JA, Scanlan KA. Spectral and color Doppler artifacts. RadioGraphics 1992; 12:35-44.
  1. Middleton WD. Color Doppler: image interpretation and optimization. Ultrasound Q 1998; 14:194-208.
  1. Wachsberg RH, Koneru B, Levine CD. Transcapsular arterial collateralization of a liver allograft after hepatic artery occlusion in an adult: color Doppler ultrasonographic diagnosis. J Ultrasound Med 1995; 14:859-861.
  1. Rosenthal SJ, Harrison LA, Baxter KG, Wetzel LH, Cox GG, Batnitzky S. Doppler ultrasound of helical flow in the portal vein. RadioGraphics 1995; 15:1103-1111