Downloaded from www.ajronline.org by 186.190.118.217 on 11/08/15 from IP address 186.190.118.217. Copyright ARRS. For personal use only; all rights reserved
505
Regenerating Nodules of Liver Cirrhosis: MR Imaging with Pathologic Correlation
i,... .
Kuni
Ohtomo1 Yuji ltai1 Yumiko Ohtomo2 Junji Shiga2 Masahiro ho1
To
.
determine
....
the
.
pathologic
..
.
basis
for
low-intensity
_________
. .
nodules
seen
on
MR
images
of
the liver in patients with cirrhosis, we obtained spin-echo and gradient-echo MR images in 20 cirrhotic patients in whom partial hepatectomies were subsequently performed for hepatocellular carcinoma. Low-intensity liver nodules were shown on gradient-echo and spin-echo images in eight patients. Pathologic study of the liver in these patients showed that these nodules were regenerating nodules containing hemosiderin. Low-intensity nodules were seen only on T2-weighted spin-echo images in four other patients. Microscopic examination of the liver in these patients showed regenerating nodules without hemosiderin deposits. Broad fibrous septa containing vascular spaces were present in two of these four patients. These results suggest that regenerating nodules containing hemosiderin or those that are surrounded by vascular fibrous septa are visible on MR images as low-intensity nodules and that gradient-echo images are useful in demonstrating nodules with hemosiderin. AJR
March
154:505-507,
1990
In a previous report, we showed that multiple, small, low-intensity nodules in the liver are often observed on MR images of patients with cirrhosis [1 ]. Some preliminary MR imaging-pathologic correlations with MR images of liver specimens ex vivo have disclosed that there is a close relationship between demonstration of low-intensity nodules on MR images and hemosiderin deposits in regenerating nodules of liver cirrhosis [2, 3]. MR imaging-pathologic correlation with in vivo MR images is necessary to confirm the results of these reports since changes in the water content and blood flow in specimens might influence the MR demonstration of low-intensity nodules. In this article, we correlated in vivo MR images of cirrhotic patients with pathologic specimens obtained after partial hepatectomies were carried out for hepatocellular carcinoma.
Materials
and Methods
Twenty
Received July 24, 1989; October 23, 1989.
accepted
after revision
consecutive
unit
(Siemens,
Department of Radiology, University of Tokyo. 7-3-1 Hongo, Bunkyo-ku, Tokyo, Japan 113. Address reprint requests to K. Ohtomo. 2 Department of Pathology. University of Tokyo,
two-dimensional
Tokyo,
obtained
Japan
113.
0361 -803X/90/1 © American
543-0505
Roentgen
Ray Society
cirrhotic
patients
with
hepatocellular
carcinoma
in whom
MR
images
were obtained within 3 months before tumor resections performed between 1986 and 1989 were included in this study. The 20 patients comprised 13 men and seven women 47-73 years old. The cause of liver cirrhosis was viral hepatitis in 19 patients (hepatitis B in six and non-A, non-B in 1 3) and schistosomiasis in one. MR imaging was performed with a Magnetom
scanning
(TR/TE) cm
Erlangen,
diameter with
50 cm,
two
addition
to SE
in all patients.
images,
with
MR images
in the remaining
(FLASH)
operated
transformation
was
and 2000/75
intervals
W. Germany)
Fourier
at 1 .5 T. Images
technique. a zooming
The matrix
factor
were
constructed
size was 256
of 1 .2 to 1 .5. Spin-echo
with
the
x 256 and the (SE)
2000/28
with a section
excitations
(17.1
seven
patients.
gradient-echo
Pulse sequences
thickness of 10 mm were obtained at 1.5mm) in 13 patients; SE 2000/22,90 images were SE 600/17 images were obtained in 12 patients. In
(GRE)
images
were
of GRE images
obtained
were
19/12
with
low-angle
shot
with a flip angle
fast
of 900
OHTOMO
506
in 19 patients
and 25/1 1/90#{176} in the other patient.
With the use of
Downloaded from www.ajronline.org by 186.190.118.217 on 11/08/15 from IP address 186.190.118.217. Copyright ARRS. For personal use only; all rights reserved
two excitations (1 2 sec), a single 1 0-mm sections as obtained during respiratory suspension. In these 20 patients, tumor resection was carried out within 3 months after MR imaging (in 1 0 patients, within 2 weeks). The size
of resected specimens ranged from 4.5 x 4.5 x 4.0 to 17.0 x 12.0 x 6.0 cm and the size of hepatocellular carcinomas from 1 .0 x 1.0 x 0.9 to 5.0 x 4.0 x 3.5 cm. Prussian routine
microscopic
was present
examinations
in regenerating
blue stain was added
to determine nodules
and
fibrous
whether
during
hemosiderin
septa.
ET AL.
AJR:154,
March
1990
rounded by thin fibrous septa in all patients, and partial thickening of the septa was seen in two of the patients in the SE group. Hemosiderin deposits in regenerating nodules were revealed only in the eight patients in the GRE/SE group; mild hemosiderin deposits also were seen in fibrous septa in two of them. Hemosiderin deposits in regenerating nodules and/ or fibrous septa were not seen in the SE and negative groups. Representative MR images and microscopic findings of the GRE/SE and SE groups are shown in Figures 1 and 2, respectively.
Results Low-intensity nodules were shown on GRE images and on SE images with a TR of 2000 msec in eight patients (the GRE/SE group) and only on SE images with a TR of 2000 msec in four patients (the SE group). In the other eight patients (the negative group), low-intensity nodules were not revealed on any MR images. In six patients in the GRE/SE group, more low-intensity nodules were seen on GRE images than on SE images and some of the nodules also were bigger on GRE images. Low-intensity nodules were shown more clearly on first-echo SE images (2000/28 or 2000/22) than on secondecho SE images (2000/75 or 2000/90) in four patients in the GRE/SE group and to the same degree in the other four patients. Low-intensity nodules were seen most clearly or only on second-echo SE images in the SE group. On Ti weighted SE images (600/1 7), low-intensity nodules were shown in three of five patients in the GRE/SE group but in none in the SE group. The size of the low-intensity nodules on MR images ranged from 3 to 15 mm in all patients in the two groups. Microscopically, the cirrhoses in these 20 patients were classified as micronodular (three), micro- and mesonodular (six), mesonodular (10), and macronodular (one). Although there was a wide variety in the size of regenerating nodules, even in single specimens, they almost always ranged from 3 to 20 mm microscopically. Regenerating nodules were sur-
Fig. 1.-Micronodular
liver cirrhosis
with hemosiderin
deposits
in regenerating
Discussion Our results suggest that a close relationship exists between MR demonstration of low-intensity nodules and pathologic findings of hemosiderin deposits in regenerating nodules. Low-intensity nodules were revealed more clearly on GRE than on SE images in the GRE/SE group, in which hemosiderin deposits in regenerating nodules were revealed microscopically. This is explained by the fact that the GRE technique including FLASH is more sensitive to the magnetic susceptibility of paramagnetic hemosiderin deposits [4]. In this study, only Ti -weighted GRE images were obtained because they were suitable for dynamic MR study of the liver tumors, which was our first concern during this period [5]; T2-weighted GRE images with long TEs and small flip angles must be more useful in demonstrating regenerating nodules with hemosiderin as low-intensity nodules because these images are more sensitive to T2 and T2* shortening owing to paramagnetic hemosiderin deposits [6]. Hemosiderin deposits in regenerating nodules have been reported to be associated with alcoholic liver injury [7], although the cause of liver cirrhosis in our patients was usually related to viral hepatitis infection. In the SE group, low-intensity nodules were revealed most clearly on T2-weighted SE images, and regenerating nodules without hemosiderin deposits were surrounded by broad
nodules
(GRE/SE
group).
A and B, Low-intensity nodules are shown more clearly on gradient-echo (GRE) image, 19/12/90#{176}(A), than on T2-weighted SE image, 2000/75 C, Hemosiderin deposits, recognized as black dots, are scattered in 3-mm regenerating nodule (arrow). (Prussian blue, original magnification
(B). x 10)
AJR:154, March 1990
REGENERATING
NODULES
OF
LIVER
CIRRHOSIS
507
If,
.
,
._.:.
Downloaded from www.ajronline.org by 186.190.118.217 on 11/08/15 from IP address 186.190.118.217. Copyright ARRS. For personal use only; all rights reserved
\.
.:
Fig. 2.-Microand mesonodular cirrhosis without hemosidenn deposits in regenerating nodules (SE group). A and B, Low-intensity nodules are shown on T2-weighted SE image, 2000/90 (A), but not on gradient-echo image, 19/12/90#{176} (B). C, Regenerating nodules are surrounded by broad vascular fibrous septa. (Heidenhain-azan, original magnification x 10)
fibrous septa in two patients. Regenerating nodules might be shown as low-intensity nodules in these two patients because broad fibrous septa that contained numerous vascular spaces had relatively high intensities on T2-weighted SE images, just like vascular scars within hepatic tumors [8]. However, in the remaining two patients in the SE group, microscopic examination could not reveal any findings that might have caused low-intensity nodules. The detectability of low-intensity nodules on Ti-weighted SE images was quite low in both groups in this study, which agreed with our previous findings [i]. Although the number of patients in this study is limited, our results suggest that regenerating nodules with hemosiderin deposits might be shown as low-intensity nodules on GRE as well as on SE images with long TRs, as is seen with sideroblastic nodules (Gamna)ofthe spleen [9, i 0]. We also suggest that there might be other pathologic entities in which regenerating nodules appearing as low-intensity nodules are shown most clearly on T2-weighted SE images; vascular fibrous septa may be one possibility. The most important limitation of this study is that direct correlation of MR findings with gross and microscopic findings was not possible because surgical specimens were usually small and ex vivo MR imaging with non-formahin-fixed surgical specimens could not be performed because of the limited availability of the MR unit. Although the size of low-intensity nodules roughly corresponded to that of regenerating nodules in surgical specimens, a more thorough direct MR imagingpathologic correlation is necessary, and MR images of fresh, whole livers obtained by autopsy or before liver transplantation are needed to confirm our preliminary results. Further
clinical experience is also necessary to determine the sensitivity of MR for demonstrating multiple low-intensity nodules on GRE and/or SE images in cirrhotic patients and to clarify the clinical significance of detecting hemosiderin deposits in regenerating nodules.
REFERENCES 1 . ltai Y, Ohnishi 5, Ohtomo K, et al. Regenerating nodules of liver cirrhosis: MR imaging. Radiology 1987; 165:419-423 2. Ohtomo K, Itai Y, Ohtomo Y, Shiga J, Minami M, ho M. MRI demonstration of pseudolobules of liver cirrhosis: particular reference to iron deposits on autopsied specimens (English abstr). Nippon Igaku Hoshasen Gakkai Zasshi 1989;49:681 -683 3. Kadoya M, Matsui 0. Yoshikawa J, et al. MRI of uneven iron deposition in the cirrhotic liver (in Japanese). Jpn J Magn Reson Med 1989:9 :269 4. Winkler ML, Thoeni RF, Luh N, Kaufman L, Margulis AF. Hepatic neoplasia: breath-holding MR imaging. Radiology 1989;170:801 -806 5. Yoshida H, Itai Y, Ohtomo K, Kokubo T, Minami M, Yashiro N. Small hepatocellular carcinoma and cavemous hemangioma: differentiation with dynamic FLASH MR imaging with Gd-DTPA. Radiology 1989;171: 339-342 6. Murakami T, Marukawa T, Sakurai K, et al. Regenerating nodules of liver cirrhosis-evaluation of gradient-echo (FLASH) (in Japanese). Jpn J Magn Reson Med 1989;9:270 7. Gal EA. Posthepatitic, postnecrotic, and nutritional cirrhosis: a pathological
analysis. Am J Pathol 1960;36:241 -271 8. Rummeny E, Weissleder R, Sironi S. et al. Central scars in primary liver tumors: MR features, specificity, and pathologic correlation. Radiology 1989:171:323-326 9. Minami M, ltai Y, Ohtomo K, et al. Sideroblastic nodules of the spleen: MR imaging of portal hypertension. Radiology 1989:172:681-684 10. Sagoh T, Itoh K, Togashi K, et al. Gamna-Gandy bodies of the spleen: evaluation with MR imaging. Radiology 1989:172:685-687