| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Gastrointestinal Imaging |
1 From the Departments of Radiology (D.Z., G.B., V.V.), Pathology (M.C., C.D.), and Hepatology (D.V.), Hopital Beaujon, Clichy, France; and Departments of Radiology (M.P.F.) and Pathology (T.W.), University of Pittsburgh Medical Center, Pa. Received January 22, 2003; revision requested April 11; revision received June 24; accepted August 8. Address correspondence to G.B., Department of Radiology, University of Palermo, Via Villaermosa 29, 90139 Palermo, Italy (e-mail: gbranca@yahoo.com).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: This was a retrospective review of congenital hepatic fibrosis cases identified at two institutions over the course of 8 years. Eight men and 10 women with an age range of 22–72 years (mean age, 39 years) were included. Contrast material–enhanced and unenhanced CT scans were obtained through the liver in all patients. Two radiologists evaluated size of and morphologic findings (atrophy or hypertrophy localized according to hepatic segments) in the liver; increased diameter or number of hepatic arteries at the hilum; presence of hepatic nodules, varices, spontaneous splenorenal shunts, and splenomegaly; and association with other hepatic ductal plate malformations and renal abnormalities.
RESULTS: Sixteen patients had morphologic abnormalities in the liver, 15 had splenomegaly (three underwent splenectomy for portal hypertension), and 14 had varices or spontaneous splenorenal shunts. An enlarged hepatic artery and a tangle of abnormally enlarged arterial vessels were identified in five and four patients, respectively, and four of these nine patients had large benign regenerative nodules. Ten patients had renal abnormalities and nine had an associated ductal plate malformation.
CONCLUSION: This retrospective study shows that certain findings (ie, liver morphologic and associated ductal plate abnormalities, varices, splenomegaly, and renal abnormalities) are frequently observed in combination in patients with congenital hepatic fibrosis.
© RSNA, 2004
Index terms: Bile ducts, abnormalities, 76.289 • Bile ducts, diseases, 76.289 • Liver, CT, 76.1211 • Liver, fibrosis, 76.289
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Because of the difficulty in the clinical and pathologic diagnosis of congenital hepatic fibrosis, imaging studies could play a crucial role in the diagnosis of this disorder if accurate and reliable signs can be determined. Unfortunately, little is known about the radiologic diagnosis of congenital hepatic fibrosis. Ernst et al (14) described the magnetic resonance (MR) cholangiographic findings in four children with congenital hepatic fibrosis. To our knowledge, however, there have been no published findings of a large series of adults with congenital hepatic fibrosis who underwent computed tomography (CT), and criteria that would allow more confident CT diagnosis have not been evaluated.
At our two referral centers for patients with hepatobiliary disease, we had the opportunity to study a group of patients with congenital hepatic fibrosis and observed certain patterns of hepatic morphologic changes and associated extrahepatic findings that seemedcharacteristic or suggestive of this entity. We hypothesized that by reviewing the CT scans of these patients we might be able to identify some imaging findings seen in this diagnosis. Thus, in this retrospective study we evaluated CT findings in adult patients with pathologically proved congenital hepatic fibrosis.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
The following widely accepted pathologic criteria (8) were used for inclusion in this study: diffuse periportal fibrosis in bands of varying thickness, irregularly shaped islands of normal hepatic tissue, numerous uniform bile ducts scattered in fibrous tissue, and paucity of the portal vein branches. Six patients were excluded because they did not meet the inclusion criteria. Reasons for exclusion at pathologic examination were insufficient material (n = 3) or no evident fibrosis and septa, together with numerous normal portal spaces (n = 3). Thus, our retrospective study included the remaining 18 patients who had complete clinical and pathologic information and CT scans.
Our study included eight men and 10 women with a mean age of 39 years. Age range was 22–72 years for men and 26–66 years for women. Three patients had a family history of congenital hepatic fibrosis. Fifteen patients had isolated or associated signs of portal hypertension, (esophageal varices [n = 10], splenomegaly or splenectomy for portal hypertension [n = 15], upper gastrointestinal bleeding [n = 5], and ascites [n = 2]). Six patients had isolated or associated signs and symptoms of cholangitis such as fever (n = 4), pain (n = 3), or jaundice (n = 2). One patient had a history of hepatitis of unknown origin since the age of 3 years. Hepatic insufficiency was diagnosed in five patients with a prothrombin time of 23.4–29.7 seconds (normal laboratory value, 11.0 seconds) and a factor V of 37%–55% (normal range, 75%–120%). Five patients had a hepatocellular injury (aspartate aminotransferase and alanine aminotransferase, >40 U/L), and 15 patients had biochemical cholestasis. All 15 patients had elevated levels of 
-glutamyltransferase (>40 U/L), and 11 had elevated levels of alkaline phosphatase (>130 U/L). Seven patients had an elevated total bilirubin level (>1.0 mg/dL [17.1 µmol/L]). No patient had normal results of a blood liver function test, although in two patients a slight elevation, 1.2 times greater than normal, of the -glutamyltransferase level was the only abnormal finding.
CT Examinations
CT examinations were performed through the liver and included contrast material–enhanced and unenhanced scanning in all patients. Three patients underwent conventional (ie, nonhelical) CT. Unenhanced scanning was followed with contrast-enhanced scanning in the portal venous phase. Partially contiguous 5-mm-thick sections were obtained with both unenhanced CT and portal venous phase CT. Helical CT was performed in 15 patients and included arterial and portal venous phase imaging (section thickness, 5–7 mm; pitch, 1.0–1.4), with delays of 25–35 seconds and 60–70 seconds, respectively, after bolus injection of intravenous contrast material. Patients received 140 mL of the intravenous iodinated contrast material ioversol (Optiray 350; Mallinckrodt Medical, St Louis, Mo) or iohexol (Omnipaque; Nycomed-Amersham, Cork, Ireland). For conventional CT, contrast material was administered at a rate of 2–3 mL/sec. For helical multiphasic examinations, the rate of injection was 4–5 mL/sec with use of a power injector (Medrad, Pittsburgh, Pa). All CT examinations were performed with DRG1 (Siemens, Erlangen, Germany), Twin Flash (Marconi Medical Systems, Cleveland, Ohio), or CT HiSpeed Advantage (GE Medical Systems, Milwaukee, Wis) scanners. For those patients who underwent more than one CT examination, only the CT scans that correlated most closely with the results of the pathologic study were reviewed.
Image Analysis
The CT scans were evaluated on hard-copy film retrospectively and jointly by two abdominal radiologists (V.V. and G.B., with 17 and 7 years of experience, respectively) who were blinded to all clinical information but who had knowledge of the diagnosis of congenital hepatic fibrosis. Disagreement was minor and resolved by consensus. On the basis of published descriptions of pathologic findings and our prior clinical observations, two radiologists worked together and analyzed the frequency of several specific findings in these patients, including liver size and morphologic characteristics (ie, liver hypertrophy or atrophy localized according to hepatic segments), dilated bile ducts and biliary calculi, the presence of signs of portal hypertension (ie, spontaneous splenorenal shunts, varices, ascites, and splenomegaly), patency of the main portal and hepatic veins, cavernous transformation of the portal vein, the size of hepatic arteries at the hilum, and the presence of hepatic nodules and their size, number, attenuation, and homogeneity.
Two radiologists working together made a subjective judgment of hepatic size relative to patient size. The sizes of the whole liver and of each segment were estimated on the CT scans, and manual mapping and three-dimensional volumetric analysis were not performed. Hepatic segments were classified as caudate, right anterior, right posterior, left medial, and left lateral. Hepatic segments were defined as hypertrophic or atrophic if the volume of individual segments was abnormally increased or reduced compared with the other hepatic segments. Since the left medial segment (segment IV) is often small in cirrhotic livers (15), we noted whether its size was normal or enlarged in our patients. The common bile duct was considered abnormally dilated if the diameter was more than 6 mm. Biliary calculi were diagnosed on unenhanced CT scans by identifying hyperattenuating foci within obviously dilated ducts.
A splenorenal shunt was defined as a spontaneous anastomosis of the splenic vein or a perisplenic varix to an enlarged left renal vein. Varices were defined as enlarged collateral veins. Spleen size was measured on transverse scans, and a splenic length of more than 12 cm was considered excessive. The portal and hepatic veins were defined as patent if the entire lumen was filled with contrast-enhanced blood. Cavernous transformation was defined as an abundance of collateral veins in the hepatic hilum that provided an alternative route around a thrombosed segment of the main portal vein or lobar branches. The hepatic artery was considered to be abnormally enlarged when its diameter was equal to or larger than that of the splenic artery or if the hepatic artery branches at the hilum were abnormally numerous. A hepatic nodule was defined as a round well-circumscribed lesion, in which attenuation was judged relative to that of the surrounding liver parenchyma on the unenhanced CT scans and on the scans obtained during all phases of contrast enhancement. A nodule was considered homogeneous if it appeared enhanced to the same degree in all its parts. One investigator (D.Z.) noted the type of pathologic diagnosis in the nodules (percutaneous or surgical biopsy, liver transplantation). For those nodules with no available pathologic diagnosis, images obtained at 1-year follow-up were evaluated.
Two radiologists working together noted the coexistence of other ductal plate malformations that have been reported to be associated with congenital hepatic fibrosis (1), including Caroli disease, von Meyenburg complexes, and choledochal cysts. Caroli disease is defined as multiple, segmental cystic dilatations of the intrahepatic bile ducts containing enhancing fibrovascular bundles (the central dot sign) (16). von Meyenburg complexes (ie, bile duct microhamartomas) were defined as small multiple lesions up to 15 mm in size and uniformly distributed throughout the liver that remained hypoattenuating but with variable enhancement after intravenous injection of contrast material (17,18). A choledochal cyst was defined as a cystic or fusiform dilatation of the common bile duct. Two radiologists working together also noted the presence of the following renal abnormalities: polycystic renal disease, parenchymal calcifications, parenchymal atrophy (a renal length, volume, or both that were less than expected for the age, sex, and height of the patients [19]) and medullary sponge kidney disease (ie, tubular ectasia), which is a fusiform or cystic dilatation of the collecting ducts in those four patients who underwent excretory urography. Two radiologists working together noted any other abnormalities or malformations in other abdominal organs. Finally, two pathologists at each institution confirmed the diagnosis of large regenerative nodules at slide review if variable-sized hepatocytes arranged in plates (one or two plates wide) and narrow sinusoids were seen.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Imaging Findings
The imaging findings in all 18 patients with congenital hepatic fibrosis are summarized in Table 1.
|
|
|
|
|
|
|
|
|
|
|
|
Portal Hypertension
Fifteen patients had signs of portal hypertension, including varices (esophageal, n = 10; other, n = 6), splenorenal shunt (n = 6) (Fig 5), ascites (n = 2), and splenomegaly or prior splenectomy (n = 15) (Fig 1). Three patients had no signs of portal hypertension.
Vascular Involvement and Nodules
Four patients had portal vein thrombosis, and one also had cavernous transformation of the portal vein. Hepatic veins were patent in all patients. An enlarged hepatic artery and a tangle of abnormally enlarged arterial vessels (Fig 4) were identified in five and four patients, respectively. Four of these nine patients had large benign regenerative nodules (Fig 3). Multiple lesions were seen in all four patients (10, five, four, and three lesions) and disseminated throughout the liver. The mean diameter of the largest benign lesion in each patient was 15 mm (range, 5–30 mm). Nodules were homogeneous in all patients. On unenhanced CT scans, all nodules were isoattenuating to normal liver and were homogeneously hyperattenuating on both hepatic arterial and portal venous phase CT scans (Fig 3).
Renal Involvement
Ten patients had the following renal abnormalities: polycystic renal disease (n = 3) and parenchymal calcifications (either associated with parenchymal atrophy [n = 4] or medullary sponge kidney disease [n = 3]).
Other Abnormalities
One patient had transmural thickening of the wall of the rectum with intramural and perirectal phleboliths that was diagnosed as a rectal hemangioma.
Table 3 shows the frequency of the individual CT findings observed in our patients for the diagnosis of congenital hepatic fibrosis. Table 4 shows the frequency of association for each of the five CT findings most commonly observed in each patient with congenital hepatic fibrosis.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
In this study, we found that some distinct CT features were frequent in congenital hepatic fibrosis, namely, liver morphologic findings, (ie, hypertrophy of left lateral segment, normal size or hypertrophy of left medial segment, and atrophy of the right lobe), varices, splenomegaly, associated ductal plate malformations, and renal abnormalities. We derived these radiologic and anatomic criteria for the diagnosis of congenital hepatic fibrosis on the basis of published descriptions of pathologic findings (1–4,6–9) and our own prior imaging experience. To our knowledge, the combination of these CT signs for the diagnosis of congenital hepatic fibrosis has not been reported previously.
Our results indicate that nine of our 18 patients (50%) had one or more associated congenital abnormalities of the biliary tree. This association is not surprising, as all these conditions belong to the same spectrum of ductal plate malformations (1). Moreover, we noted coexisting renal abnormalities in 10 patients. The association of congenital hepatic fibrosis with renal abnormalities is in accordance with the findings of earlier studies (2,7,14,22). Furthermore, congenital hepatic fibrosis has been described in association with numerous other organ abnormalities (1), and we report a case of associated hemangioma of the rectum. The association of congenital hepatic fibrosis with abnormalities in other viscera or systems could be explained by simultaneous aberrant embryonic development of multiple organs and systems (1).
In our study, we noted an enlarged hepatic artery or a tangle of abnormally numerous arterial vessels at the liver hilum. Desmet (1) has made similar observations. Moreover, in our series four patients had a total of 22 hyperattenuating hepatic lesions, which were proved to be large multiacinar regenerative nodules. Large regenerative nodules have been found in association with other causes of portal hypertension and vascular derangement of the liver (23), particularly Budd-Chiari syndrome (24), and are frequently seen in patients with a larger than normal hepatic artery (24). Some authors (24–26) believe that large regenerative nodules are the consequence of augmented arterialization of the liver. Our findings of large multiacinar regenerative nodules in patients with congenital hepatic fibrosis and enlarged hepatic arteries are a further clue to this association.
Several investigators have reported that quantitative analysis of the regional changes in hepatic morphology on cross sectional images is useful in the diagnosis of liver disease. In a study of 36 patients with primary sclerosing cholangitis, Dodd et al (27) reported that lobulation of the liver contour, atrophy of the lateral or posterior hepatic segments, and hypertrophy of the caudate lobe occurred more frequently in patients with primary sclerosing cholangitis induced end-stage cirrhosis than in patients with end-stage cirrhosis of other causes. Blachar et al (28) reported that CT findings in primary biliary cirrhosis include generalized or segmental hypertrophy prior to generalized or segmental atrophy in the terminal stage of the disease. Lafortune et al (15) compared the size of the medial segment in a group of cirrhotic patients with the medial segment in a control group and found a substantial reduction in the size of the medial segment in cirrhotic patients.
In our patients with congenital hepatic fibrosis, we found some distinctive hepatic morphologic findings (along with associated abnormalities). While atrophy of the right lobe and hypertrophy of the left lateral segment and the caudate lobe are common observations that are also seen in patients with advanced viral or alcoholic cirrhosis (29), we were surprised to note that only one patient had a small medial segment, which was otherwise of normal or enlarged size in the remaining 17 patients. We believe that this morphologic finding may be useful in distinguishing patients with congenital hepatic fibrosis from those with cirrhosis.
It is important to recognize the limitations of our study. Because this was a retrospective review, selection bias was unavoidable, although case selection was based on pathologic rather than imaging criteria. Because pathologic findings of congenital hepatic fibrosis are considered specific but not sensitive, and because we required pathologic proof in all cases, it is likely that we excluded some patients from analysis. Moreover, while congenital hepatic fibrosis is usually diagnosed in patients who are 5–20 years of age, we limited our search to two adult hospitals, and this is reflected in the higher mean age of our patients (39 years) compared with the mean age (22 years) reported in another series (6). Congenital hepatic fibrosis can remain clinically silent for many years (3–8,30), however, which makes it important for physicians who care for adult patients to be aware of this diagnosis.
We may have underreported associated renal abnormalities, especially renal tubular ectasia, because we relied primarily on CT diagnoses, and only four patients underwent excretory urography. Of the four patients with benign regenerative nodules, two underwent biopsy of one nodule only, which left a total of six nodules without pathologic proof. While we cannot be certain that these were benign regenerative nodules as well, these were all smaller than 1 cm and had the same imaging characteristics at multiphasic helical CT as the nodules with biopsy proof. Moreover, 1-year follow-up data in one patient showed stability in size of the nodules. Last, we did not compare our series of patients to a control group, such as cirrhotic patients, that could have had similar findings; therefore, further research is needed to assess the specificity of the signs we describe.
In conclusion, our results show that certain findings (liver morphologic findings, associated ductal plate abnormalities, varices, splenomegaly, and renal abnormalities) are frequently observed in combination in patients with congenital hepatic fibrosis and, therefore, can be helpful in making a diagnosis.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. Vilgrain, B. Condat, C. Bureau, A. Hakime, A. Plessier, D. Cazals-Hatem, and D. C. Valla Atrophy-Hypertrophy Complex in Patients with Cavernous Transformation of the Portal Vein: CT Evaluation Radiology, October 1, 2006; 241(1): 149 - 155. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Brancatelli, M. P. Federle, V. Vilgrain, M.-P. Vullierme, D. Marin, and R. Lagalla Fibropolycystic Liver Disease: CT and MR Imaging Findings RadioGraphics, May 1, 2005; 25(3): 659 - 670. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
