AJR 2004; 182:459-462
© American Roentgen Ray Society
T2-Shortening Effect of Fibrinogen Inclusions on MRI of Hepatocellular Carcinoma: Case Report and Experimental Relaxation Measurement
Makoto Fujita1,2,
Takashi Horinouchi1,
Shingo Ishiguro3,
Ryu Ishihara4,
Hiroshi Kasugai4,
Terumasa Yamada5,
Yo Sasaki5,
Hiroshi Maeda1,
Etsuo Inoue1 and
Chikazumi Kuroda1
1 Department of Diagnostic Radiology, Osaka Medical Center for Cancer and
Cardiovascular Disease, 3 Nakamichi, 1-chome, Higashinari-ku, 537-8511 Osaka,
Japan.
2 Department of Radiology, Nishinomiya Municipal Central Hospital, 8-24
Hayashida-cho, Nishinomiya, 663-8014 Hyogo, Japan.
3 Department of Pathology, Osaka Medical Center for Cancer and Cardiovascular
Disease, Osaka, Japan.
4 Department of Gastroenterology, Osaka Medical Center for Cancer and
Cardiovascular Disease, Osaka, Japan.
5 Department of Surgery, Osaka Medical Center for Cancer and Cardiovascular
Disease, Osaka, Japan.
Received January 6, 2003;
accepted after revision July 8, 2003.
Address correspondence to M. Fujita
(m.fujita{at}minos.ocn.ne.jp).
Introduction
Moderately differentiated hepatocellular carcinoma generally shows high
signal intensity relative to adjacent hepatic parenchyma on T2-weighted images
[1]. We describe a rare case of
moderately differentiated scirrhous hepatocellular carcinoma that showed
markedly low signal intensity on T2-weighted images. A histologic feature of
the tumor was abundant intracytoplasmic fibrinogen inclusions (pale bodies)
[2,
3]. To examine the possible
effects of fibrinogen on signal intensity on T2-weighted images, we performed
a phantom study in which experimental relaxation measurements were made of
fibrinogen solutions. The T2-shortening effect of fibrinogen and its clinical
significance as a possible cause of low signal intensity on T2-weighted images
are discussed.
Case Report
A 70-year-old woman with chronic hepatitis was found to have a low-density
area 3 cm in diameter in the right hepatic lobe on unenhanced CT. MRI was
performed with a 1.5-T scanner (Magnetom Vision, Siemens, Erlangen, Germany).
The lesion showed signal intensity slightly lower than that of the surrounding
liver tissue on T1-weighted fast spin-echo images. However, on T2-weighted
fast spin-echo images with fat suppression (TR/TE, 2,800/128; echo-train
length, 23), signal intensity was markedly lower in most of the lesion, except
for a central stellate area that showed slightly higher signal intensity
(Fig. 1A). Dynamic fast
low-angle shot (FLASH) imaging (135/4.1; flip angle, 90°) was performed
with gadopentetate dimeglumine (0.1 mmol/kg of body weight). On early-phase
images, most of the lesion showed marked enhancement that had washed out on
late-phase images. In contrast, the central area of the lesion was minimally
enhanced on early-phase images and was gradually enhanced on late-phase images
(Fig. 1B). Hepatic
arteriography and subsequent CT arteriography showed the same hemodynamics as
had dynamic MRI.
View larger version (149K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —70-year-old woman with scirrhous hepatocellular carcinoma.
Transverse T2-weighted image with fat suppression (TR/TE, 2,800/128;
echo-train length, 23) shows 3-cm-diameter lesion in right hepatic lobe. Most
of mass is markedly hypointense relative to remaining liver, except for
central hyperintense area.
|
|
View larger version (130K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —70-year-old woman with scirrhous hepatocellular carcinoma.
Arterial phase transverse T1-weighted fast low angle shot image (135/4.1; flip
angle, 90°) in dynamic MRI with gadopentetate dimeglumine reveals obvious
enhancement in most of tumor. However, minimal enhancement is shown in central
area of tumor.
|
|
Because the hypervascularity of the lesion was compatible with moderately
differentiated hepatocellular carcinoma, partial hepatectomy was performed. On
gross examination, the resected specimen showed a well-defined, lobulated
tumor with a central fibrous scar. Microscopic examination showed that the
tumor consisted of neoplastic hepatocytes with a trabecular pattern. Cell
cords were separated by fibrous bands. Fibrosis was more prominent in the
central area of the tumor. The histologic diagnosis was moderately
differentiated hepatocellular carcinoma with sclerotic change. The most
striking feature of the tumor was the large number of intracytoplasmic
inclusion bodies. These inclusions contained a pale eosinophilic substance
consistent with pale bodies, which are composed of fibrinogen in the dilated
rough endoplasmic reticulum (Fig.
1C). These inclusions were less prevalent in the central area of
the tumor. The Prussian blue reaction was carried out to investigate the
possible effect of iron content in the tumor. The entire tumor was unstained
(not shown).
View larger version (196K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C. —70-year-old woman with scirrhous hepatocellular carcinoma.
Photomicrograph of resected specimen shows hepatocellular carcinoma cell cords
separated by fibrous bands. Large number of pale, eosinophilic inclusion
bodies (pale bodies) (arrow) are seen in tumor cells. (H and E,
x200)
|
|
Experimental relaxation measurements of fibrinogen solutions were taken
using the method of Gore et al.
[4] to examine the effects of
fibrinogen on T2-weighted imaging. To determine the approximate tissue
concentration of fibrinogen in our patient, we hypothesized that the maximal
tissue concentration of fibrinogen corresponds to the area ratio of pale
bodies on a histologic field of view. The area ratio of fibrinogen on a
histologic slide (Fig. 1C) was
calculated with an image analyzer (MacScope, Mitani, Fukui, Japan). Because
the area ratio was 9.6%, experimental solutions of human fibrinogen (Beriplast
P, Aventist Parma, Frankfurt, Germany) were prepared in concentrations of 0%,
2%, 4%, 6%, 8%, and 12% in saline. We also compared the relaxation efficacy of
fibrinogen to albumin as a control macromolecule. Solutions of human albumin
(Kenketsu Albumin, Yoshitomi, Osaka, Japan) in saline were prepared in the
same range of concentrations. Twelve tubes filled with different
concentrations of fibrinogen or albumin were set up in a 4-mmol solution of
copper sulfate. MRI of the test solutions (phantom) was performed with a
clinical scanner (1.5-T Magnetom Vision) at room temperature (24°C). A
program loaded in the MR scanner was used to calculate relaxation times. T1
was measured with a spin-echo sequence (TRs = 250, 1,000, 1,500, 2,000, 3,000,
and 5,000 msec; TE = 15 msec), and T2 was measured with the
Carr-Purcell-Meiboom-Gill sequence with 16 echoes (TR = 9,000 msec; TE =
50–800 msec). The same measurements were repeated after 180°
conversion of the z-axis direction. Data were estimated using
nonlinear least-squares two-parameter fits.
Figure 2A shows an
experimental T2-weighted image (9,000/200) of the phantom. Signal intensity
decreased markedly with increasing fibrinogen concentration. Fibrinogen had a
greater effect on signal intensity than did albumin.
Figure 2B shows the T2
calculation image of the phantom. Fibrinogen had a T2-shortening effect
greater than that of albumin. Figure
2C shows the 1 / T2 relaxation rates for both fibrinogen and
albumin as functions of concentration. Linear regression analysis gave the
following relations:
where f is the percentage concentration of fibrinogen and a
is the percentage concentration of albumin. The T1/T2 ratios for fibrinogen
and albumin at the higher concentration were approximately 14.2 and 4.6,
respectively.
View larger version (116K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A. —Experimental MRI of phantom and measured relaxation rates.
Coronal T2-weighted spin-echo image (TR/TE, 9,000/200) of phantom shows marked
decrease in signal intensity with increasing fibrinogen concentration (lower
two tube lines: concentration increases from right of fourth line to left of
third line). Effect of fibrinogen is more intense than that of albumin (upper
two lines: concentration increases from right of second line to left of first
line).
|
|
View larger version (119K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. —Experimental MRI of phantom and measured relaxation rates.
Coronal T2-calculated image of phantom shows that fibrinogen has T2-shortening
effect greater than that of albumin (arrangement of tubes is same as described
in A).
|
|
Discussion
Early-stage well-differentiated hepatocellular carcinoma occasionally shows
low signal intensity relative to surrounding hepatic parenchyma on T2-weighted
images [1]. However, moderately
differentiated hypervascular hepatocellular carcinoma rarely shows low signal
intensity on T2-weighted images
[1].
Mature fibrosis with collagenous tissue has low signal intensity on
T2-weighted images, whereas immature or early fibrosis with numerous
fibroblasts may have high signal intensity on T2-weighted images
[5]. On T2-weighted images,
hepatocellular carcinoma with sclerotic change, even the fibrolamellar
variant, generally has high signal intensity, except for a central scar with
low signal intensity [6]. In
our patient, both histologic and hemodynamic studies showed more prominent
fibrosis in the central area of the tumor. These findings suggest that
fibrosis cannot be responsible for the low signal intensity on T2-weighted
images in our case. A possible effect of iron content on signal intensity was
also ruled out by results of the Prussian blue reaction.
Pale bodies are intracytoplasmic inclusions occasionally seen in
hepatocellular carcinoma that represent accumulation of fibrinogen in dilated
rough endoplasmic reticulum [2,
3]. A certain correlation
between pale bodies and sclerotic change of hepatocellular carcinoma tissue
was suggested. Tumor cells in our patient showed a large number of pale
bodies; the distribution of the pale bodies within the tumor corresponds well
with the area of low signal intensity on T2-weighted images. Fibrinogen is a
macromolecule with a molecular weight of 400,000, which is several times
greater than that of albumin. These findings suggest that fibrinogen affects
signal intensity on T2-weighted images.
There have been few studies of the effects of fibrinogen on MRI
[7]. Stuhlmuller et al.
[7] studied the effect of
varying fibrinogen and hematocrit concentrations on MR relaxation times of
thrombus at fibrinogen concentrations of 1, 10, and 100 µmol/L
(nearly 4%) and concluded that fibrinogen has a T2-shortening effect on
thrombus at lower hematocrit values. The aim of this study was to examine the
general effect of fibrinogen at clinical tissue concentrations on MR
relaxation using a standard method of relaxometry. Thus, we used a
conventional spin-echo sequence in the experimental study instead of the fast
spin-echo sequence used in the patient study. However, a possible difference
between these two sequences is one of the limitations of this study because a
long echo-train length generally results in lower T2 signal intensity.
Our experimental relaxation measurements suggest that fibrinogen has an
intense T2-shortening effect greater than that of albumin, which may be
dependent on molecular weight
[8]. The T1/T2 ratio in
fibrinogen shows that the T2-shortening effect is stronger than the
T1-shortening effect, in accordance with the lack of an obvious elevation of
signal intensity on T1-weighted images. Another potential limitation of our
study is a difference in the effect on relaxation times between extracellular
fibrinogen and intracellular fibrinogen, as is present in the hepatocellular
carcinoma cells.
In conclusion, our findings suggest that fibrinogen is a possible cause of
low signal intensity on T2-weighted imaging.
Acknowledgments
We thank Masao Okazaki, The Jikei University School of Medicine, for his
help in manuscript preparation.
References
- Ebara M, Fukuda H, Kojima Y, et al. Small hepatocellular carcinoma:
relationship of signal intensity to histopathologic findings and metal content
of the tumor and surrounding hepatic parenchyma.
Radiology1999; 210:81
–88[Abstract/Free Full Text]
- Nakashima O, Sugihara S, Eguchi A, Taguchi J, Watanabe J, Kojiro M.
Pathomorphologic study of pale bodies in hepatocellular carcinoma.
Acta Pathol Jpn1992; 42:414
–418[Medline]
- Craig JR, Peters RL, Edmondson HA, Omata M. Fibrolamellar carcinoma
of the liver. Cancer1980; 46:372
–379[Medline]
- Gore JC, Brown MS, Mizumoto CT, Armitage IM. Influence of glycogen
on water proton relaxation times. Magn Reson Med1986; 3:463
–466[Medline]
- Lee JKT, Glazer HS. Controversy in the MR imaging appearance of
fibrosis. Radiology 1990;177
: 21–22[Free Full Text]
- Ichikawa T, Federle MP, Grazioli L, Madariaga J, Nalesnik M, Marsh
W. Fibrolamellar hepatocellular carcinoma: imaging and pathologic findings in
31 recent cases. Radiology1999; 213:352
–361[Abstract/Free Full Text]
- Stuhlmuller JE, Olson JD, Burns TL, Skorton DJ. Effect of varying
fibrinogen and hematocrit concentrations on magnetic resonance relaxation
times of thrombus. Invest Radiol1992; 27:341
–345[Medline]
- Gore JC, Kennan RP. Physical and physiological basis of magnetic
relaxation. In: Higgins CB, Hricak H, Helms CA, eds. Magnetic
resonance imaging of the body, 2nd ed. New York, NY: Raven Press,1992
: 33–42
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Introduction
Case Report
