DOI:10.2214/AJR.04.1695
AJR 2006; 186:168-173
© American Roentgen Ray Society
Contrast-Enhancing Hepatic Eosinophilic Abscess During the Hepatic Arterial Phase: A Mimic of Hepatocellular Carcinoma
Jae Ho Byun1,
Dong Hyun Yang1,
Seong Eon Yoon1,
Hyung Jin Won1,
Yong Moon Shin1,
Yong Yeon Jeong2 and
Se Jin Jang3
1 Department of Radiology, Asan Medical Center, University of Ulsan College of
Medicine, 388-1 Pungnap-2dong, Songpagu, Seoul 138-736, Korea.
2 Department of Diagnostic Radiology, Chonnam National University Medical
School, Gwang-Ju, Korea.
3 Department of Pathology, Asan Medical Center, University of Ulsan College of
Medicine, Seoul, Korea.
Received November 1, 2004;
accepted after revision December 22, 2004.
Address correspondence to J. H. Byun
(jhbyun{at}amc.seoul.kr).
Abstract
OBJECTIVE. Our objective was to review CT, MR, angiographic, and
sonographic images obtained in patients with histopathologically proven
contrast-enhancing hepatic eosinophilic abscesses during the hepatic arterial
phase.
CONCLUSION. On CT, MR, and angiographic images, eosinophilic abscess
in the liver can show contrast enhancement during the hepatic arterial phase.
If patients have peripheral hypereosinophilia or are at low risk for
hepatocellular carcinoma or hepatic metastasis, those with small
arterial-enhancing hepatic nodules need further preoperative evaluation, such
as a sonographically guided biopsy.
Keywords: conventional angiography • dynamic CT • liver disease • MRI
Introduction
Eosinophilic abscess of the liver is caused by various disorders including
allergic disorder, parasitic infestations, neoplastic disease, and
hypereosinophilic syndrome
[1-3].
The imaging findings of eosinophilic abscess have been reported in the
literature
[1-6].
On CT, eosinophilic abscesses are usually multiple, small (usually < 2 cm),
subtly hypoattenuating, and round or oval lesions with a fuzzy margin, and are
most conspicuous on the portal phase. The usual distribution of eosinophilic
abscesses is multifocal subcapsular or periportal. On sonographic examination,
these abscesses usually appear as small, poorly defined, and round or oval low
echoic nodules.
In daily clinical practice, eosinophilic abscess with typical or atypical
imaging findings can mimic metastasis in patients with underlying
gastrointestinal or hepatic malignancy
[3,
7,
8].
We report an uncommon imaging finding of eosinophilic abscess of the
liver.
Materials and Methods
We reviewed the imaging findings of three patients with arterial enhancing
eosinophilic abscess proven by histopathology. All patients were men (age
range, 43-54 years). Two patients were transferred to our hospital because of
an incidentally found hepatic nodule on ultrasonography at another hospital.
The other patient underwent sonography for a checkup at our hospital and a
hepatic nodule was found. Three patients were asymptomatic. The viral markers
for hepatitis B and C were negative in all three patients. One patient had
acute transverse myelitis and the others had no underlying disease and no
history of allergy or medication use. No patient had an identified cause of
hepatic eosinophilic abscess. The peripheral eosinophilic count was 0.2%,
10.2%, and 12.5%, respectively (normal range, 1-10% of WBCs). Histopathologic
confirmation was obtained by segmentectomy in two patients and by
sonographically guided core biopsy in the other patient. The interval between
histopathologic confirmation and radiologic examination was 12, 16, and 27
days, respectively. CT was performed on two patients, MRI on two, sonography
on one, and angiography on one.
In our hospital, the follow-up period for the three patients after
histopathologic confirmation was 4, 12, and 26 months, respectively. In two
patients, the results of follow-up CT and the peripheral eosinophilic count
were unremarkable. The other patient who received sonographically guided core
biopsy had only a follow-up laboratory test. That patient's last peripheral
eosinophilic count was 12.2%.
Our institutional review board does not require its approval or a patient's
informed consent for this type of retrospective review.
We evaluated the size, shape, margin, and enhancement pattern of the
eosinophilic abscesses on CT and MRI and the echogenicity, size, and shape on
sonography. We also analyzed the clinical and laboratory findings including
peripheral eosinophilic count and underlying disease.
CT was performed using an MDCT scanner (LightSpeed QX/i, GE Healthcare). CT
scans were obtained during the hepatic arterial phase (using a bolus tracking
technique or a 36-sec delay) and the portal venous phase (using a 72-sec
delay) after IV injection of 150 mL of iopromide (Ultravist 370; Schering) at
a rate of 3 mL/sec with a 5-mm section thickness, a pitch of 3:1 (high-quality
mode), and 5-mm reconstruction intervals during the hepatic arterial phase and
portal venous phase. MRI was performed on a 1.5-T unit (Gyroscan Intera,
Philips Medical Systems). The sequences were axial T1-weighted spoiled
gradient echo (TR/TE, 155/4.6; flip angle, 80°) followed by axial
T2-weighted turbo spin-echo (1600/80; 90° flip angle). Axial
T2* spoiled gradient-echo MR images (142/9.2; flip angle, 18°)
were also acquired 10 min after IV bolus injection of 1.4 mL of ferucarbotran
(Resovist; Schering) followed by a 20-mL saline flush. In another patient,
echo-fast gradient-echo 3D images (7-9/2-3; 15° flip angle) with spectral
fat saturation were obtained after administration of gadopentetate dimeglumine
(Magnevist; Berlex). After acquiring unenhanced images, 0.1 mmol/kg of
contrast material was injected IV as a rapid bolus and was followed by 10 mL
of normal saline flush. The images were obtained at 30-35 sec (hepatic
arterial phase), 70-90 sec (portal venous phase), and 3-5 min (delayed phase)
after contrast injection. Section thickness was 5 mm; field of view, 350 mm;
and matrix, 256 x 256-512 x 512. Sonography was performed using an
Acuson Sequoia 512 scanner (Siemens Medical Solutions) with 3.5-5.0-MHz convex
transducers. Sonographically guided percutaneous biopsy was performed with an
18-gauge automated biopsy gun. For angiography, arterial vascular access was
obtained with femoral arterial punctures using the Seldinger technique. An
angiographic catheter (Yashiro, Terumo) was selectively placed; one was placed
in the superior mesenteric artery for indirect portogram and the other in the
common hepatic artery for hepatic arteriogram.
View larger version (103K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —53-year-old man with arterial-enhancing eosinophilic abscess
proven by segmentectomy in segment VII of liver. Axial CT scan during hepatic
arterial phase shows round enhancing nodule with ill-defined margin
(arrows).
|
|
View larger version (122K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —53-year-old man with arterial-enhancing eosinophilic abscess
proven by segmentectomy in segment VII of liver. Corresponding axial portal
venous phase CT scan shows round hypoattenuating nodule with ill-defined
margin (arrows).
|
|
Results
The imaging findings of eosinophilic abscess in the three patients are
summarized in Table 1. The size
of the eosinophilic abscess was 7, 10, and 13 mm, respectively. Its shape was
ovoid in two patients and round in the other patient. Its margin was ill
defined in all patients. On CT, the eosinophilic abscess was isoattenuating
with liver parenchyma on unenhanced CT scan in two cases. It was
hyperattenuating with liver parenchyma during the hepatic arterial phase and
then hypoattenuating during the portal venous phase in two cases (Figs.
1A,
1B,
1C,
1D,
1E,
1F,
3A,
3B, and
3C).
View larger version (156K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D —53-year-old man with arterial-enhancing eosinophilic abscess
proven by segmentectomy in segment VII of liver. Corresponding axial
T2-weighted MR image shows nodule as hyperintense at same site
(arrows).
|
|
View larger version (146K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1E —53-year-old man with arterial-enhancing eosinophilic abscess
proven by segmentectomy in segment VII of liver. Corresponding axial
ferucarbotran-enhanced T2* MR image shows hyperintense nodule
(arrows).
|
|
View larger version (184K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1F —53-year-old man with arterial-enhancing eosinophilic abscess
proven by segmentectomy in segment VII of liver. Photomicrograph of specimen
obtained at segmentectomy shows inflammatory cell infiltrates predominantly
composed of eosinophils. (H and E, x100)
|
|
View larger version (147K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —54-year-old man with arterial-enhancing eosinophilic abscess
proven by biopsy in segment V of liver. Axial CT scan during the hepatic
artery phase shows subtly enhancing nodule with very ill-defined margin
(arrows).
|
|
View larger version (156K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B —54-year-old man with arterial-enhancing eosinophilic abscess
proven by biopsy in segment V of liver. Corresponding axial portal venous
phase CT scan shows small hypoattenuating nodule at same site
(arrows).
|
|
View larger version (102K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3C —54-year-old man with arterial-enhancing eosinophilic abscess
proven by biopsy in segment V of liver. Intercostal sonogram of liver shows
subtly hypoechoic nodule with somewhat well-defined margin
(arrows).
|
|
On MRI, low signal intensity was observed on T1-weighted images and high
signal intensity on T2-weighted images in two patients (Figs.
1A,
1B,
1C,
1D,
1E, and
1F). Gadolinium-enhanced
hepatic arterial phase MRI showed heterogeneous enhancement, and
gadolinium-enhanced portal venous phase MRI showed heterogeneous low signal
intensity (Figs. 2A,
2B, and
2C). Ferucarbotran-enhanced
T2*-weighted MRI showed high signal intensity (Figs.
1A,
1B,
1C,
1D,
1E, and
1F). Ferucarbotran-enhanced
T1-weighted MRI also showed high signal intensity. Hepatic arteriogram showed
a small round, subtle enhancing nodule (Figs.
2A,
2B, and
2C).
View larger version (139K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —43-year-old man with arterial-enhancing eosinophilic abscess
proven by segmentectomy in segment VI of liver. Anteroposterior view of
hepatic arteriogram shows small nodular subtle enhancement with ill-defined
margin (arrows).
|
|
View larger version (121K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —43-year-old man with arterial-enhancing eosinophilic abscess
proven by segmentectomy in segment VI of liver. Corresponding axial
gadolinium-enhanced T1-weighted MR image during hepatic arterial phase shows
small nodule with ill-defined margin as subtly heterogeneously hyperintense
(arrows).
|
|
View larger version (98K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C —43-year-old man with arterial-enhancing eosinophilic abscess
proven by segmentectomy in segment VI of liver. Corresponding axial
gadolinium-enhanced T1-weighted MR image during portal venous phase shows
small nodule with ill-defined margin as subtly heterogeneously hypointense
(arrows).
|
|
Sonographic examination showed a somewhat well-marginated, round,
heterogeneous low echoic nodule without posterior shadowing or enhancement
(Figs. 3A,
3B, and
3C).
Before histopathologic examination, we presumed three lesions were
hepatocellular carcinoma. The histopathologic diagnosis was eosinophilic
abscess in all three patients.
Discussion
In many literature reports
[1-6],
several terms, such as "foci of eosinophil-related necrosis,"
"focal eosinophilic infiltration," "focal eosinophilic
necrosis," and "eosinophilic abscess" have been used for
eosinophil-related hepatic lesions. These terms have not yet been clearly
defined histopathologically. In general, eosinophilic abscess means
eosinophilic infiltration with focal necrosis, and focal eosinophilic
infiltration means eosinophilic infiltration without focal necrosis.
The imaging findings of eosinophilic abscess or focal eosinophilic
infiltration are well known and are usually similar regardless of the cause.
On CT, eosinophilic abscess or focal eosinophilic infiltration usually appears
as small, oval or round, and low attenuating nodules with well- or ill-defined
margins. They are most conspicuous during the portal venous phase. They have
iso/low, low/low, or mixed attenuation on the hepatic arterial/portal venous
phases
[1-3,
5]. On sonographic examination,
they manifest as small (usually < 2 cm), poorly or well-defined, oval or
round, low echoic nodules without hyper- or hypoechoic rims
[1,
2,
4]. There are two case reports
to our knowledge of MRI of focal eosinophilic infiltration
[7,
8]. In these reports, MRI
showed a nodule with hyperintensity on T2-weighted images and isointensity on
T1-weighted images. On ferumoxide-enhanced imaging, there were high and
isointense signals on T2-weighted images and T1-weighted images, respectively
[8].
Yoo et al. [2] reported that
in one of 20 patients, two of six foci of eosinophilic infiltration showed
high attenuation on hepatic artery phase CT and low attenuation on portal
venous phase CT. Two case reports
[7,
8] showed five cases of focal
eosinophilic infiltration that were combined with hepatocellular carcinoma or
gastric cancer, and appeared as homogeneous hyperattenuation on CT hepatic
arteriography, hypoattenuation on CT arterial portography, and as a nodule
with subtle enhancement on gadolinium-enhanced arterial phase MRI. Over 40
months, enhancing eosinophilic abscess of the liver occurred in three of 203
cases of hepatic eosinophilic abscess in our hospital.
On histopathology, eosinophilic abscess or focal eosinophilic infiltration
is related to perivascular infiltration of eosinophils, predominantly in the
periportal space, and hepatocellular necrosis
[1,
3,
5,
9,
10]. We can speculate that
portal flow obstruction by eosinophilic infiltration makes focal eosinophilic
infiltration or eosinophilic abscess hypoattenuation on portal phase CT, and
inflammatory hyperemia or compensatory arterial overflow causes enhancement on
arterial phase CT and MRI. However, the transient hepatic attenuation
difference associated with portal flow cessation shows transient
hyperattenuation during the hepatic arterial phase, and then the
hyperattenuating area becomes isoattenuating or slightly hyperattenuating with
liver parenchyma during the portal venous phase. The enhancement of an
eosinophilic abscess or focal eosinophilic infiltration during the hepatic
arterial phase did not have a straight border or a wedge shape that is usually
seen in the transient hepatic attenuation difference. We will need further
study to clarify this difference between eosinophilic abscess and the
transient hepatic attenuation difference.
Although we performed many radiologic examinations, we could not
differentiate an arterial-enhancing eosinophilic abscess or focal eosinophilic
infiltration from hepatocellular carcinoma or hypervascular hepatic
metastasis. For exact diagnosis of an arterial-enhancing nodule in patients at
low risk for primary or metastatic hepatic malignancy, special attention must
be given to the clinical and laboratory findings. Tissue diagnosis is also
necessary in these patients.
In summary, eosinophilic abscess or focal eosinophilic infiltration of the
liver rarely shows contrast enhancement during the hepatic arterial phase on
various radiologic imaging techniques. If patients have peripheral
hypereosinophilia or are at low risk for hepatocellular carcinoma or hepatic
metastasis, those who have small arterial-enhancing hepatic nodules warrant
further evaluation.
Acknowledgments
The authors thank Ah Young Kim, Pyo Nyun Kim, Hyun Kwon Ha, Moon-Gyu Lee,
and Bonnie Hami for their assistance in preparing the manuscript.
References
- Lee WJ, Lim HK, Lim JH, Kim SH, Choi SH, Lee SJ. Foci of
eosinophil-related necrosis in the liver: imaging findings and correlation
with eosinophilia. AJR 1999;172
: 1255-1261[Abstract/Free Full Text]
- Yoo SY, Han JK, Kim YH, Kim TK, Choi BI, Han MC. Focal eosinophilic
infiltration in the liver: radiologic findings and clinical course.
Abdom Imaging 2003;28
: 326-332[CrossRef][Medline]
- Jang HJ, Lee WJ, Lee SJ, Kim SH, Lim HK, Lim JH. Focal eosinophilic
necrosis of the liver in patients with underlying gastric or colorectal
cancer: CT differentiation from metastasis. Korean J
Radiol 2002; 3:240
-244[Medline]
- Kim GB, Kwon JH, Kang DS. Hypereosinophilic syndrome: imaging
findings in patients with hepatic involvement. AJR1993; 161:577
-580[Abstract/Free Full Text]
- Lim JH, Lee WJ, Lee DH, Nam KJ. Hypereosinophilic syndrome: CT
findings in patients with hepatic lobar or segmental involvement.
Korean J Radiol 2000;1
: 98-103[Medline]
- Cho DH, Chang JC, Seong KH. Dynamic CT findings of eosinophilic
hepatic abscess and its clinical course [in Korean]. J Korean
Radiol Soc 1996; 35:899
-904
- Won JH, Kim MJ, Kim BM, et al. Focal eosinophilic infiltration of
the liver: a mimick of hepatic metastasis. Abdom
Imaging 1999; 24:369
-372[CrossRef][Medline]
- Kwak HS, Han YM, Lee JM. Focal eosinophilic infiltration of the
liver mimicking heaptocellular carcinoma. Clin Imag2004; 28:36
-39[CrossRef][Medline]
- Foong A, Scholes JV, Gleich GJ, Kephart G, Holt PR.
Eosinophil-induced chronic active hepatitis in the idiopathic
hypereosinophilic syndrome. Hepatology1991; 13:1090
-1094[CrossRef][Medline]
- Croffy BC, Kopelman RK, Kaplan M. Hypereosinophilic syndrome:
association with chronic active hepatitis. Dig Dis Sci1988; 33:233
-239[CrossRef][Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
