| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Report |
1
Department of Radiology, University of Pittsburgh Medical Center,
UPMC-Presbyterian, 200 Lothrop St., Pittsburgh, PA 15213.
2
Department of Radiology, University of Brescia, Italy.
3
Department of Radiology, University of Bologna, Italy.
4
Department of Radiology, University of Pisa, Italy.
Received June 18, 2001;
accepted after revision August 31, 2001.
Address correspondence to M. P. Federle.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. Large regenerative nodules are a characteristic feature of Budd-Chiari syndrome and other hepatic vascular disorders. CT and MR imaging show consistent features of the nodules and the surrounding liver that may allow distinction of Budd-Chiari nodules from other types of hypervascular hyperplastic or dysplastic nodules.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Large regenerative nodules are known to occur in response to abnormal perfusion of the liver [2,3,4], most commonly with Budd-Chiari syndrome [5]. We report the CT and MR imaging findings in 14 patients with proven large regenerative nodules, who had a variety of hepatic perfusion abnormalities.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Study Population
Our study population was composed of nine women, one girl, and four men,
ranging from 12 to 67 years old (mean age, 39 years). Seven women had
Budd-Chiari syndrome, caused by a hypercoagulable state (n = 4) or of
unknown cause (n = 3). A summary of the other clinical and imaging
findings is presented in Table
1. No patients had viral or alcoholic hepatitis.
|
Histopathology and Follow-Up
The pathology records were reviewed retrospectively by two investigators.
The diagnosis of large regenerative nodules was based on histopathologic
analysis of specimens in all patients obtained by means of percutaneous core
needle (n = 8) or surgical (n = 2) biopsy or liver resection
(n = 1) in 11 patients; the entire liver was available in three
patients with Budd-Chiari syndrome who had liver transplantation.
Histopathologic confirmation of large regenerative nodules was based on the
findings of variable size hepatocytes arranged in plates (one or two plates
wide) and narrow sinusoids. In three patients, the gross and microscopic
features of the resected livers were analyzed, including the number and size
of the lesions and the presence and arrangement of hepatocytes, portal tracts,
and bile ducts.
For the 11 patients who did not undergo liver transplantation, it was
considered impractical to biopsy all of the hepatic nodules. Lesions were
considered to represent large regenerative nodules if they were similar to
biopsy-proven large regenerative nodules in the same liver in size and imaging
characteristics (CT or MR imaging) and if the lesions remained stable in size
over time with normal levels of serum 
-fetoprotein. All patients have
had follow-up from 12 to 60 months.
Imaging Protocols
CT was performed through the liver in 13 patients and MR imaging was
performed through the liver in seven patients. CT was performed with a
conventional transverse scanner (9800 Advantage; General Electric Medical
Systems, Milwaukee, WI) or a helical scanner (HiSpeed Advantage or LightSpeed
QX/i, General Electric Medical Systems; Somatom Plus 4, Siemens, Erlangen,
Germany). The nonhelical scan (n = 1) was obtained after a monophasic
injection of IV contrast material at 2.5 mL/sec for a total volume of 150 mL.
The helical CT examinations (n = 12) were performed with unenhanced
(n = 11), hepatic arterial dominant phase (n = 12), and
portovenous dominant phase (n = 12) images. The arterial phase series
was obtained with a scan delay of 25-35 sec, and the portal phase at 60-70 sec
after initiation of the IV bolus of contrast material. All patients received
either iothalamate meglumine (Conray 60; Mallinckrodt Medical, St. Louis, MO)
or ioversol (Optiray 350; Mallinckrodt Medical) injected IV at a rate of 3-5
mL/sec with a power injector (OP 100; Medrad, Pittsburgh, PA).
Six patients underwent MR imaging with a T1-weighted spin-echo sequence (TR range/TE range, 140-700/12-20) and seven patients underwent a T2-weighted spin-echo sequence (4000-9230/70-140; echo-train length, 8 or 16). T1-weighted imaging was repeated in all patients after IV contrast material administration. Patients received gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) or gadobenate dimeglumine (Multihance; Bracco, Milan, Italy) at a dose of 0.1 mmol/kg of body weight, and underwent biphasic MR imaging during the hepatic arterial phase (25- to 35-sec delay) and portovenous phase (60- to 70-sec delay). Three patients were imaged 3 hr after the IV administration of gadobenate dimeglumine. MR imaging studies were performed with Signa 1.5-T imagers (General Electric Medical Systems), or a Symphony 1.5-T imager (Siemens).
Image Analysis
The CT scans were reviewed retrospectively and jointly by three abdominal
radiologists who had knowledge of the diagnosis of large regenerative nodules
but were not aware of the specific number of tumors or clinicopathologic
findings in any patient. Data on interobserver and intraobserver variability
were not calculated because the sample size was small and because our goal was
not to assess the accuracy of cross-sectional imaging in the diagnosis of
large regenerative nodules, but to describe the findings and to perform a
correlation with the pathology findings. We used majority opinion, and all
observers agreed on the number of masses in all cases.
The lesion characteristics sought on CT and MR images included the number, size, and attenuation or intensity. A nodule was defined as a well-circumscribed round lesion. The attenuation and signal characteristics of the lesions were judged relative to those of the surrounding liver parenchyma on the unnenhanced CT and MR images, as well as on the images obtained during all phases of contrast enhancement (hepatic arterial, portovenous, and delayed phase). A nodule was considered homogeneous if it enhanced to the same degree in all its parts. When a patient had serial studies, the size stability of a lesion was assessed. The presence of a surgically created shunt was noted.
Institutional review board approval was obtained for this study. Our review board does not require patient consent for this type of study and none was obtained.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
No hepatocellular carcinoma was identified, and the levels of
-fetoprotein were normal in all patients (
20 µg/L).
Imaging
More than 20 nodules were seen in each of eight patients; in the remaining
six patients, all three reviewers detected one lesion in each of three
patients, three lesions in one patient, five lesions in one patient, and 12
lesions in one patient. The size of the nodules was 0.5-4 cm.
On unenhanced CT, all nodules were isoattenuating to normal liver, except in four patients who had diffuse low attenuation caused by steatosis (n = 1) or Budd-Chiari syndrome (n = 3), resulting in hyperattenuating nodules. All nodules were hyperattenuating on hepatic arterial and portovenous phases (Figs. 1A,1B and 2). Six patients (three with more than 20 lesions) had several nodules that showed hypoattenuating rings (Fig. 3A,3B,3C,3D,3E).
|
|
|
|
|
|
|
|
On MR imaging in seven patients, all had nodules that were hyperintense on T1-weighted images, and three had some isointense nodules. All nodules showed intense homogeneous enhancement with IV bolus gadolinium, but fewer than half showed a hypointense ring (Figs. 3A,3B,3C,3D,3E and 4A,4B,4C) and only one patient showed a hyperintense ring (Fig. 3A,3B,3C,3D,3E). Approximately 10% of the nodules returned to isointensity on the portovenous phase. Three patients, one with more than 20 lesions and two with one lesion each, underwent delayed imaging 3 hr after administration of gadobenate dimeglumine; all of the nodules showed retained hyperintensity, some with peripheral enhancement (Fig. 4A,4B,4C).
|
|
|
On T2-weighted imaging, few nodules were seen, and these were isointense or hypointense in all but one patient, in whom there was one hyperintense nodule (Figs. 3A,3B,3C,3D,3E and 4A,4B,4C).
In none of the nodules was there CT or MR imaging evidence of fat, necrosis, or calcification.
All but two patients have had follow-up from 12 to 60 months; none had a
rise in the serum -fetoprotein during this period, and only one had an
increase in the number of nodules.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Other radiologists have used the term nodular regenerative hyperplasia to describe the benign hepatic nodules that they detected by CT or MR imaging in patients with Budd-Chiari syndrome. This term may be appropriate for these lesions when present in a liver that is not fibrotic or cirrhotic, but the lesions in nodular regenerative hyperplasia are generally only about 1 mm in diameter, occasionally with clusters of lesions up to 10 mm in diameter centered on a large portal tract. Because the lesions we describe are larger and because we had histologic evidence of hepatic fibrosis between the nodules in some of our patients, the term large regenerative nodules or multiacinar regenerative nodules is more accurate.
The large regenerative nodules that can be recognized as discrete lesions on CT or MR imaging constitute only a low percentage of all regenerative nodules that occur in patients with hepatic vascular disorders, probably because most of these lesions are small. At the coordinating institution, only six of 120 patients who had proven nodular regenerative hyperplasia and biphasic helical CT had focal hypervascular lesions recognized on CT.
We think that the morphologic and hemodynamic information provided by CT and MR imaging provides important insights into the nature of focal hepatocellular nodules and the surrounding liver. The large regenerative nodules that we report were almost all multiple (Fig. 1A,1B), isodense to normal unenhanced liver, uniformly hypervascular, and hyperattenuating on both arterial and portovenous CT scans (Figs. 1A,1B and 4A,4B,4C). They varied in size from 0.5 to 4 cm. MR imaging also reflected the hypervascularity of the lesions: they were hyperintense on T1-weighted images (Figs. 3A,3B,3C,3D,3E and 4A,4B,4C) and isointense or hypointense on T2-weighted images. The three patients studied with gadobenate dimeglumine showed uptake and retention of enhancement in all of the nodules, reflecting hepatocellular composition and delayed clearance of bile from the nodules (Fig. 4A,4B,4C).
All except one of the 14 patients in our series had conditions that caused abnormal hepatic perfusion and that are known to be associated with nodular hyperplasia of the liver [2]. Seven patients had Budd-Chiari syndrome, and the others had various conditions that altered hepatic blood flow, such as restrictive cardiomyopathy, congenital absence of the portal vein (Fig. 4A,4B,4C), and idiopathic superior mesenteric arteriovenous fistula. Wanless [4] has concluded that large regenerative nodules are the result of heterogeneous distribution of hepatic blood flow or chronic ischemia. Large nodules are known to occur in livers with hepatic or portal vein obliteration, and they can develop before clinical signs of cirrhosis or portal hypertension [3]. The histopathologic features of these large nodules correlate well with imaging features reported here and by other investigators [5]. The nodules have an increased arterial supply corresponding to their bright enhancement on CT and MR imaging. They often have atrophic tissue in their periphery or in surrounding liver with curvilinear areas of sinusoidal dilatation and marked congestion, accounting for the perinodular rings that were hypoattenuating on bolus-enhanced CT and T1-weighted MR images that we observed in some of our patients (Figs. 3A,3B,3C,3D,3E and 4A,4B,4C). The only hyperintense ring that we observed might have been caused by recent hemorrhage or congestion (Fig. 3A,3B,3C,3D,3E). Large regenerative nodules are drained by hepatic veins, making them prone to congestion and infarction if venous drainage is impaired, as it is in Budd-Chiari syndrome or cardiomyopathy. Infarcted regenerative nodules in cirrhosis are described by Kim et al. [6] as hyperintense at T2-weighted pulse sequences. Infarction can therefore be an explanation for the few cases of hyperintense regenerative nodules on T2-weighted imaging described by some authors [5] as well as for the only case we observed.
Wanless [4] has found mineral deposits in these nodules, which is in accordance with our findings and perhaps accounts for the hyperintensity of some of the nodules on T1-weighted imaging (Fig. 2) that we and others [5] have reported.
Pathologists have reported metaplastic bile ductules in some large regenerative nodules [4], which we also confirmed in our patients. This would account for the delayed clearance (prolonged enhancement) that we observed in our three patients studied with gadobenate dimeglumine, a hepatobiliary MR contrast agent.
Hepatocellular carcinoma is detected by conventional CT in only about 50% of cases. Helical multiphasic CT has increased the sensitivity of detection of hepatocellular carcinoma in both the cirrhotic liver [7] and the noncirrhotic liver. Characteristically, hepatocellular carcinoma is hypoattenuating to liver on unenhanced and delayed CT; is hyperattenuating on hepatic arterial dominant phase CT; and may be hypo-, iso-, or hyperattenuating on portovenous dominant phase CT. Other useful characteristics of hepatocellular carcinoma are heterogeneity, multiplicity, lesions of different sizes, encapsulation, and portovenous or hepatovenous invasion. Similar features are evident on MR imaging, and hepatocellular carcinoma is usually hypointense to liver on T1-weighted imaging and hyperintense on T2-weighted imaging.
We believe it is important for radiologists and referring physicians to recognize large regenerative nodules and, in particular, to distinguish these from hepatocellular carcinoma. Before we became familiar with this entity, several of the patients we describe here were being followed as if they had multifocal hepatocellular carcinoma and, as a result, could have been denied potentially curative therapy (transplantation) or offered inappropriately aggressive therapy (chemotherapy).
It might be argued that Budd-Chiari syndrome and other vascular disorders that lead to large regenerative nodules might also give rise to hepatocellular carcinoma or that these nodules might themselves be considered premalignant; however, there is little or no evidence to support this argument. All of our patients with Budd-Chiari syndrome either underwent liver transplantation (with detailed analysis of all nodules in the explanted liver) or had long-term follow-up to prove stability of the nodular lesions. We performed a recent literature review and found reports of fewer than 10 patients who had a diagnosis of both Budd-Chiari syndrome and hepatocellular carcinoma. Several of these patients had coexisting chronic viral hepatitis, leading the authors to speculate that the viral infection, rather than the hepatic vascular occlusion, was the causative factor for hepatocellular carcinoma [8]. Other authors [5] reported patients who had a simultaneous diagnosis of Budd-Chiari syndrome and hepatocellular carcinoma, leading to speculation that the hepatocellular carcinoma caused the vascular compromise rather than its having resulted from Budd-Chiari syndrome.
In conclusion, large regenerative nodules are hyperplastic hypervascular lesions that are caused by vascular derangement of the liver due to decreased portovenous or hepatovenous flow. The resulting multifocal hepatic arterial dilatation leads to the focal hyperplasia and proliferation of hepatocytes that we recognize as large regenerating nodules, assuming adequate size of the lesions and imaging techniques that capture arterial phase enhancement of the liver. Although these nodules have some imaging features that overlap with some other types of hyperplastic and dysplastic nodules, we believe that the combination of features found on CT or MR imaging along with clinical and imaging evaluation of the liver should allow confident diagnosis and treatment in most cases.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Cura, Z. Haskal, and J. Lopera Diagnostic and Interventional Radiology for Budd-Chiari Syndrome RadioGraphics, May 1, 2009; 29(3): 669 - 681. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Kandpal, R. Sharma, S. Gamangatti, D. N. Srivastava, and S. Vashisht Imaging the Inferior Vena Cava: A Road Less Traveled RadioGraphics, May 1, 2008; 28(3): 669 - 689. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. F. Hanna, D. A. Aguirre, N. Kased, S. C. Emery, M. R. Peterson, and C. B. Sirlin Cirrhosis-associated Hepatocellular Nodules: Correlation of Histopathologic and MR Imaging Features RadioGraphics, May 1, 2008; 28(3): 747 - 769. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. H Kiesewetter, N. Sheron, J. J Vettukattill, N. Hacking, B. Stedman, H. Millward-Sadler, M. Haw, R. Cope, A. P Salmon, M. C Sivaprakasam, et al. Hepatic changes in the failing Fontan circulation Heart, May 1, 2007; 93(5): 579 - 584. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Brancatelli, V. Vilgrain, M. P. Federle, A. Hakime, R. Lagalla, R. Iannaccone, and D. Valla Budd-Chiari Syndrome: Spectrum of Imaging Findings Am. J. Roentgenol., February 1, 2007; 188(2): W168 - W176. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Qayyum, A. Graser, A. Westphalen, R. B. Merriman, L. D. Ferrell, B. M. Yeh, and F. V. Coakley CT of Benign Hypervascular Liver Nodules in Autoimmune Hepatitis Am. J. Roentgenol., December 1, 2004; 183(6): 1573 - 1576. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Zeitoun, G. Brancatelli, M. Colombat, M. P. Federle, D. Valla, T. Wu, C. Degott, and V. Vilgrain Congenital Hepatic Fibrosis: CT Findings in 18 Adults Radiology, April 1, 2004; 231(1): 109 - 116. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. A. Krinsky and G. Israel Nondysplastic Nodules That Are Hyperintense on T1-Weighted Gradient-Echo MR Imaging: Frequency in Cirrhotic Patients Undergoing Transplantation Am. J. Roentgenol., April 1, 2003; 180(4): 1023 - 1027. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Brancatelli, M. P. Federle, L. Grazioli, R. Golfieri, and R. Lencioni Benign Regenerative Nodules in Budd-Chiari Syndrome and Other Vascular Disorders of the Liver: Radiologic-Pathologic and Clinical Correlation RadioGraphics, July 1, 2002; 22(4): 847 - 862. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
