| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Clinical Observations |
1 Division of Radiology, Cleveland Clinic Foundation, 9500 Euclid Ave., A21,
Cleveland, OH 44195.
2 Department of General Surgery, Cleveland Clinic Foundation, Cleveland,
OH.
Received June 28, 2005;
accepted after revision September 4, 2005.
Address correspondence to E. M. Remer.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. Intrahepatic portal venous shunts are uncommon hepatic vascular anomalies that are often not associated with manifestations of liver disease or symptoms. They are most often solitary and in the left hepatic lobe. Identification of 25 intrahepatic portal venous shunts at a single institution over 6 years suggests that they may be more common than previously known and that with an increasing use of imaging, they may be identified more often in the future.
Keywords: CT • hepatobiliary imaging • intrahepatic portal venous shunts • intrahepatic portosystemic venous shunts • liver • shunts
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The circumstances surrounding the discovery of an intrahepatic portal venous shunt in a patient without cirrhosis are variable. Shunts may be discovered at the time of commonly performed imaging studies in patients without signs or symptoms. A few patients may present with hepatic encephalopathy due to high-output shunting. In this situation, hepatic dysfunction prompts an imaging examination and, thus, discovery of the shunt. In a series in Japan of noncirrhotic patients with portal systemic encephalopathy, 36.2% of 47 patients presenting with encephalopathy had intrahepatic portosystemic shunts [3]. Hepatic encephalopathy has been more commonly reported in Japan than in the United States.
The prevalence of an intrahepatic portal venous shunt as an incidental finding in asymptomatic patients is unknown. With the increased use of imaging studies, these lesions are likely to be increasingly encountered. To our knowledge, prior reports have included between one and five patients [4-18], and no nonangiographic case series exists. We report the imaging features of incidentally discovered asymptomatic intrahepatic portal venous shunts.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
The medical record of each patient was evaluated for evidence of potential symptoms related to the shunt. None of the 22 patients manifested hepatic encephalopathy or right upper quadrant abdominal pain.
Imaging Studies
Seventeen patients were imaged with helical CT only, two with helical CT
and MRI, one with sonography, one with sonography and CT, and one with all
three techniques. CT was performed on Somatom Plus 4, Plus 4 Volume Zoom, or
Sensation 16 scanners (Siemens Medical Solutions). Scanning was performed at
2.5-, 3-, or 5-mm collimation, 2.5- to 3-mm slice interval with 2.5- or 5-mm
slice thickness. Scanning was performed in the portal venous phase only in 17
patients; before administration of contrast material and during the arterial
and portal venous phases in three patients; and before administration of
contrast material and during the arterial and delayed phases in one patient.
Patients received IV 150 mL of iodinated contrast material (iopromide,
Ultravist, Berlex; 300 mg I/mL). Contrast material was administered at a rate
of 2-4 mL/s.
Three patients were imaged with a total of five sonographic examinations. Both gray-scale and color Doppler imaging were performed. Three patients were imaged with a total of six MR examinations on a 1.5-T scanner (Symphony, Siemens Medical Solutions) with a phased-array body coil. Axial in- and opposed-phase T1-weighted 2D gradient-echo FLASH, fast spin-echo T2-weighted (turbo spin-echo), and unenhanced and enhanced 3D gradient-echo VIBE (volumetric interpolated breath-hold examination) sequences were performed in all patients. Gadopentetate dimeglumine (Magnevist, Berlex; 0.1 mmol/kg) was administered IV with a power injector, and a portal venous phase 3D gradient-echo sequence was performed. In one patient, coronal and oblique coronal half-Fourier single-shot turbo spin-echo (HASTE) images were obtained.
Evaluation and Analysis
All data, including clinical data, radiologic findings, and clinical
outcomes, were collected retrospectively. One reviewer retrospectively
reviewed images. An intrahepatic portal venous shunt was confirmed when direct
communication between a portal vein and a hepatic vein was visible and was
isoattenuating or isointense to the hepatic veins on CT or MRI, respectively.
On sonography, Doppler examination showed the venous communication with a
venous flow pattern. The following information was noted: type of draining
vein (hepatic vein, inferior vena cava); location of communication based on
Couinaud-Bismuth nomenclature
[19]; number of shunts in each
patient; whether an aneurysm was present between the portal and systemic veins
(focal enlargement between the two veins) and whether it was > 1 cm or <
1 cm; and whether the portal vein diameter was < 2 cm or > 2 cm. Also
noted were the presence of nonintrahepatic portal venous shunt hepatic lesions
and whether morphologic indicators of cirrhosis (such as an enlarged caudate
lobe and lateral segment of the left lobe, surface nodularity, or prominence
of fissures caused by atrophy
[20]) were present.
Intrahepatic portal venous shunts were classified by the system described by Park et al. [21]. Type I is a single large tubular vessel of constant diameter that connects the right portal vein to the inferior vena cava; type II is a peripheral shunt in which single or multiple communications are found between peripheral branches of portal and hepatic veins in one hepatic segment; type III is an aneurysmal communication between the peripheral portal and hepatic veins; and type IV is multiple, diffuse communications between peripheral portal and hepatic veins in both lobes of the liver.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Nineteen patients had a solitary intrahepatic portal venous shunt, 15 in the left lobe and four in the right lobe. Three patients had two intrahepatic portal venous shunts (3/22, 13.6%): Two had one in each lobe and one had two in the left lobe.
Using the classification of Park et al. [21], one shunt was type I, 11 were type II (11/25, 44%), and 13 were type III (13/25, 52%) (Fig. 3A, 3B). Of the three patients with two shunts each, each had one type II and one type III. Of the type III shunts (those with an aneurysmal structure connecting the portal and hepatic veins), three were < 1 cm and 10 were > 1 cm.
|
|
|
|
|
In two patients, the portal vein measured > 2 cm. No focal extrahepatic portal vein aneurysms were seen.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The origin of these shunts is the matter of differing opinions. When a portal vein-hepatic vein communication is seen in a patient without liver disease or a history of trauma, it is presumed to be spontaneous or congenital in origin [24-26]. Some authors have speculated that these shunts represent persistent embryonic venous anastomoses [24, 25, 27] caused by the failure of regression of connections among tributaries of the vitelline vein (the precursor of the portal and hepatic veins and portions of the inferior vena cava). Others suggest that rupture of a portal vein aneurysm into the hepatic vein is the cause [27].
That 19 (76%) of 25 of shunts in our series occurred in the left lobe seems to support the congenital theory. During development, the right umbilical vein involutes and the left umbilical vein forms a direct communication with the ductus venosus (right hepatocardiac channel), bypassing the sinusoidal plexus of the liver [28, 29]. Blood therefore flows from the placenta through the umbilical vein, ductus venosus, into the right hepatocardiac channel (later part of the inferior vena cava). After birth, the left umbilical vein forms the ligamentum teres and the sinus venosus forms the ligamentum venosum. Both the ligamentum teres and the ligamentum venosum are contiguous to the left hepatic lobe. Possibly these shunts represent persistent developmental communications.
Twenty-five shunts being incidental findings on imaging studies over a 6-year period at a single institution, albeit in approximately 282,000 studies, suggests that such shunts may well be more common than previously thought. We speculate that they will be increasingly encountered with the ever-increasing use of imaging studies and the increased spatial resolution afforded by technical advances such as MDCT and 3D, near-isotropic MRI sequences.
Controversy exists in the literature as to the clinical importance of intrahepatic portal venous shunts. In general terms, hepatic encephalopathy consists of some impairment of liver function [30]—usually a chronic liver disease—that can be aggravated by portal-systemic venous shunting. Although some authors believe that intrahepatic portal venous shunts alone may cause encephalopathy [2, 3] and that this may be worse in older patients [3, 31], this theory does not fit the current understanding of the pathophysiology of hepatic encephalopathy.
Other pathophysiologic changes that may have clinical relevance relate to the degree of shunting. A large fistula can theoretically have adverse systemic hemodynamic effects or even cardiac failure with the development of a hyperdynamic systemic circulation. The potential of long-standing diversion of portal flow away from the sinusoids to have an adverse effect on the liver—such as the development of fibrosis—has been suggested [32]. The small size of most intrahepatic portal venous shunts makes it unlikely that they have a significant clinical effect.
Several systems have been developed to classify intrahepatic portosystemic shunts. Park et al. [21] categorized shunts into four morphologic types and found that in 14 cases reported in the literature by 1990, a single large tube of constant diameter connecting the right portal vein to the inferior vena cava was the most common type. This type of lesion has been thought to occur more commonly in patients with cirrhosis [2] and was present in only one patient in our series. Tanoue et al. [2] report that of 50 shunts reported in 2003, 35 (70%) had aneurysmal communications. These aneurysmal communications accounted for 54% of the shunts in our series.
Two of 16 shunts that had interval follow-up for nonhepatic malignancy showed an increase in size. One initially had a direct tubular portal-hepatic vein communication and on subsequent imaging had evidence of an aneurysmal communication. Neither of these patients had clinically significant events related to the fistulas. Three patients in our series had two intrahepatic portal venous shunts (3/22, 13.6%), which is fewer than the 25% of patients (2/8) in the series by Tanoue et al. [2] of symptomatic patients who had more than one shunt. No direct sign or symptom from the shunts, such as right upper quadrant pain or hepatic encephalopathy, was identified in any of our patients.
The most common indication for imaging patients in our series was the presence of a nonhepatic primary malignancy in 45.8% of our patients. The prospective interpretations of studies in this series distinguished metastases in two patients and hemangiomas in an additional two patients with shunts. Other authors have cautioned that the aneurysmal component in some shunts might have the appearance of a simple cyst on gray-scale sonography or might appear as a highly and homogeneously enhancing mass on contrast-enhanced CT [31]. Cursory evaluation of CT or MRI of this type of shunt might lead to an incorrect diagnosis of a hypervascular liver lesion. Unfortunately, we were unable to determine in our retrospective series the number of shunts that might have been misdiagnosed as metastases or other focal liver lesions. However, high-resolution CT or MR images or color Doppler sonograms should provide visualization of the feeding portal vein and draining hepatic vein.
Our study has several limitations. First, we had no gold standard confirmation that the abnormalities identified as intrahepatic portal venous shunts in this series were truly these lesions. We suspect that because these patients were asymptomatic and the shunts were incidental to the indication for imaging, no additional confirmatory studies such as angiography were done. Second, our retrospective methodology introduces selection bias because only lesions described in the prospective interpretations could be found by our search. False-negative interpretations may have led to patients who had intrahepatic portal venous shunts being missed; therefore, no comment can be made about the prevalence of the lesion in our population. Third, our radiology information system search may have missed some shunts because consistent terminology was not used in reporting these findings.
In conclusion, intrahepatic portal venous shunts have been thought to be rare, described mostly in case reports or small series. Finding 25 of these at a single institution over a 5.5-year period suggests that they are may be more common than previously recognized. Most shunts in our series were in the left lobe and had an aneurysmal communication between the portal and hepatic veins. Radiologists studying patients with liver disorders should be aware of this vascular anomaly and should also recognize that many occur in asymptomatic patients without liver disease and as such do not require treatment.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
