Vous êtes sur la page 1sur 7

JOURNAL OF MAGNETIC RESONANCE IMAGING 18:189 –195 (2003)

Original Research

Hepatocellular Carcinoma of Diffuse Type: MR Imaging Findings and Clinical Manifestations

Masayuki Kanematsu, MD, Richard C. Semelka, MD, * Polytimi Leonardou, MD, Maria Mastropasqua, MD, and Joseph K.T. Lee, MD

Purpose: To assess MR imaging findings and clinical man- ifestations of diffuse-type hepatocellular carcinoma (HCC).

Materials and Methods: We retrospectively reviewed our experience with diffuse HCC from November 1994 to Octo- ber 2001. MR imaging findings and clinical features were assessed.

Results: Twenty-two consecutive patients with diffuse-type HCC (19 men and three women, age range 16 –80 years [mean, 52 years]) were identified in a review of liver MR stud- ies. This represented 13% of all patients with HCC imaged during this time period. Diffuse HCC showed a permeative, infiltrative pattern with ill-defined borders and no evidence of convex margination in all cases. At least 50% of the liver volume was involved with tumor. Diffuse-type HCC showed hypointensity in 15 patients, mixed intensity in three, and isointensity in four on T1-weighted images; heterogeneous hyperintensity in 16 patients; and homogeneous hyperinten- sity in six on T2-weighted MR images. Diffuse-type HCC showed patchy enhancement in 12 patients, miliary enhance- ment in nine, and minimal enhancement in one on postcon- trast early-phase images, and showed heterogeneous wash- out in all patients on postcontrast late-phase images. Proximal portal venous tumor thrombosis was seen in all patients. Serum -fetoprotein (AFP) value was elevated ( 10 ng/mL) in 14 of 18 patients, and 13 showed a value greater than 500 ng/mL. The four patients who did not have elevated AFP had tumors which were indistinguishable from those in patients with elevated AFP; they also did not have a distinctive clinical history.

Conclusion: Diffuse-type HCC was typically seen as an extensive, heterogeneous permeative hepatic tumor, with portal venous tumor thrombosis on MR images in all cases. Early enhancement, observed as patchy in 12 and miliary in nine of 22 patients, was a distinctive imaging feature. Elevated serum AFP value was a common finding; however, 22% had normal values.

Department of Radiology, University of North Carolina, Chapel Hill, North Carolina.

Current address (M.K.): Department of Radiology, Gifu University School of Medicine, Gifu, Japan.

*Address reprint requests to: R.C.S., Department of Radiology, CB 7510, University of North Carolina, Chapel Hill, NC 27599-7510. E-mail: richsem@med.unc.edu

Received February 4, 2003; Accepted April 8, 2003. DOI 10.1002/jmri.10336 Published online in Wiley InterScience (www.interscience.wiley.com).

Key Words: Magnetic resonance; hepatocellular carcino- ma; diffuse type; cirrhosis; contrast enhancement

J. Magn. Reson. Imaging 2003;18:189 –195. © 2003 Wiley-Liss, Inc.

HEPATOCELLULAR CARCINOMA (HCC) is the most common primary malignant neoplasm arising in pa- tients with chronic liver damage, and is most often related to hepatitis virus infection (1,2), alcohol abuse (3), or iron overload (4,5). A 1984 pathologic report (6) described that the common major gross patterns were expanding, spreading, and multifocal by means of ag- gregating a total of 529 HCC cases from Japan, the United States, and South Africa. Further modifications have divided HCC into infiltrative or expansive, single or multinodular, and mixed types, using observations on encapsulation and intrahepatic venous spread (7). Pathologically, diffuse-type HCC has been considered as a tumor that spreads throughout most of the liver and is not recognized as a focal tumor, typically accom- panied by extensive portal venous tumor thrombosis and substantial elevation of serum -fetoprotein (AFP) value (8). Although many reports have described the MR appearance of focal forms of HCC, there is a relative paucity of descriptions of the MR appearance of diffuse- type HCC in the literature. Diffuse-type HCC is often difficult to detect on imaging studies because of its permeative appearance and heterogeneity of back- ground chronic liver disease. The purpose of this study was to describe the MR imaging findings and clinical manifestations of diffuse-type HCC.

MATERIALS AND METHODS

We retrospectively searched the radiologic records of MR imaging of the liver performed at the Department of Radiology, University of North Carolina, from November 1994 to October 2001, and found records of 171 pa- tients who had HCC and underwent MR imaging of the liver. Among this group, 22 patients were considered to have diffuse-type HCC based on the criterion that there was no distinct margination on any portion of the tumor on any sequence. Only patients with no prior treatment were included to avoid misinterpreting treatment changes as findings of tumor. Fourteen patients, in- cluding four with normal serum AFP values and four in

© 2003 Wiley-Liss, Inc.

189

190 Kanematsu et al.
190
Kanematsu et al.

Figure 1. A 51-year-old man with diffuse-type HCC in alcoholic cirrhosis, showing serum alpha-fetoprotein value of 38000 ng/mL. A: Unenhanced T1-weighted spoiled gradient-echo (140/4.1) axial MR image shows the tumor (arrows) involving the entire right lobe of the liver as areas of homogeneous, mild hypointensity. Note a small volume of perihepatic and perisplenic ascites (curved arrow). B: Unenhanced T2-weighted fat-suppressed spin-echo (2400/90) axial MR image shows the tumor as areas of heterogeneous, moderate hyperintensity (arrows). C: Gadolinium-enhanced early-phase spoiled gradient-echo (140/ 4.1) axial MR image shows the tumor as areas of patchy enhancement (arrow). D: Gadolinium-enhanced fat-suppressed late-phase spoiled gradient-echo (140/4.1) axial MR image shows the tumor as areas of multiple foci of wash-out (arrow). Portal venous tumor thrombosis in the left main portal vein branch is shown as areas of discrete hypointensity (curved arrow).

whom serum AFP values were not available, had his- topathologic confirmation (laparotomic wedge biopsy [N 1], percutaneous needle-core aspiration biopsy [N 13]). The remaining eight patients without histo- logic confirmation had MR imaging findings consistent with extensive, malignant hepatic tumors associated with markedly high ( 2,000 ng/mL) serum AFP values. MR imaging was performed with a 1.5-T MR imager (Vision, Siemens Medical Systems, Iselin, NJ). MR im- aging included a T1-weighted, in-phase, breath-hold, spoiled gradient-echo sequence (repetition time of 140 – 175 msec and echo time of 4.1–4.5 msec [140 –175/ 4.1–4.5]; flip angle, 80°; section thickness, 8 mm; in- tersection gap, 20%; one signal acquired; 21 sections in a 20-second breath hold), and a T2-weighted sequence performed on fat-suppressed spin-echo (2400/90; sec- tion thickness, 8 mm; intersection gap, 20%; two sig- nals acquired) (one patient), short tau inversion recov- ery turbo spin-echo (5,110/76/170 [TR/TE/TI]; section thickness, 8 mm; intersection gap, 20%; two signals acquired) (four patients), or half-Fourier turbo spin- echo sequence ( /90; section thickness, 8 mm; inter-

section gap, 20%; one signal; 20 sections) (17 patients). Transverse spoiled gradient-echo images were acquired prior to and after intravenous bolus injection of 0.1 mmol/kg gadolinium chelate (Magnevist, Berlex Labo- ratories, Wayne, NJ or Omniscan, Nycomed, New York, NY). Postcontrast spoiled gradient-echo sequences were initiated at 18 seconds and one minute, and fat-sup- pressed spoiled gradient-echo sequence was acquired two minutes postcontrast. Three investigators, Masa Kanematsu, Richard C. Se- melka, and Polytimi Leonardou, who were blinded to the clinical and histopathologic information and to the original MR imaging reports, retrospectively evaluated, in consensus, the transverse unenhanced T1- and T2- weighted MR and postcontrast early- and late-phase MR images in each patient. The MR images were re- viewed for the following findings: tumor extension in the liver expressed as number of Couinaud’s segments in- volved by tumors, fibrous capsules or septa, signal in- tensity characteristics of tumors on unenhanced T1- and T2-weighted MR images, contrast enhancement characteristics of tumors on early- and late-phase post-

MRI of Diffuse HCC 191
MRI of Diffuse HCC
191

Figure 2. A 44-year-old man with diffuse-type HCC in cirrhosis due to type-C viral hepatitis, showing serum alpha-fetoprotein value of 9000 ng/mL. A: Unenhanced T1-weighted spoiled gradient-echo (170/4.1) axial MR image shows the tumor (arrow) involving the right hepatic lobe and the medial segment of the left hepatic lobe as ill-demarcated areas of homogeneous, mild hypointensity, accompanied by extensive portal venous tumor thrombosis (curved arrow) shown as areas of mixed intensity. Note a moderate amount of perihepatic and perisplenic ascites (small arrows). B: Unenhanced T2-weighted turbo spin-echo ( /90) axial MR image shows the tumor (arrow) as areas of heterogeneous, moderate hyperintensity. Portal venous thrombosis (curved arrow) is shown as areas of mixed signal intensity. C: Gadolinium-enhanced early-phase spoiled gradient-echo (170/4.1) axial MR image shows the tumor as areas of miliary enhancement (arrow). Portal venous tumor thrombosis is shown as areas of heterogeneous hypointensity (curved arrow). Note areas of necrosis in the tumor (small arrow). D: Gadolinium-enhanced late-phase fat-suppressed spoiled gradient-echo (175/4.1) axial MR image shows the tumor as areas of heterogeneous isoin- tensity with some areas of wash-out (arrow). Portal venous tumor thrombosis is shown as areas of discrete, heterogeneous, hypointensity (curved arrow).

contrast MR images, bile duct dilatation, portal venous tumor thrombosis, hepatic venous tumor thrombosis, ascites, and upper abdominal lymph-node metastases. The signal intensity of tumor on T1-weighted images was categorized as moderately hypointense when it was comparable to that of the spleen, and mildly hypoin- tense when it was intermediate between liver and spleen. The signal intensity of tumor on T2-weighted images was categorized as moderately hyperintense when it was comparable to spleen, and mildly hyperin- tense when it was intermediate between liver and spleen. Enhancement characteristics of tumor were categorized into patchy and miliary: patchy enhance- ment referred to foci of enhancement that were irregu- lar in size and shape and ill-defined, and miliary en- hancement referred to foci that were small ( 5 mm) and relatively well defined. Other information on underlying hepatic disease (vi- ral hepatitis, alcohol abuse, autoimmune hepatitis,

hemochromatosis, etc.), serological laboratory test re- sults (total bilirubin, AFP, alkaline phosphatase, ala- nine aminotransferase, aspartate aminotransferase, -glutamyltranspeptidase), and distant metastases were obtained using the clinical information system of our institution.

RESULTS

From November 1994 to October 2001, we had 22 pa- tients with diffuse-type HCC (Figs. 1–3), which repre- sented 13% of all patients with HCC. The 22 patients included 19 men and three women, ranging in age from 16 –80 years (mean age, 52 years). Sixteen patients had cirrhosis, and the remaining six had chronic hepatitis. The patient characteristics are summarized in Table 1 and the MR imaging findings in Table 2. Serum total bilirubin value (normal range, 0 –1.2 mg/ dL) was elevated in 19 of 20 patients in whom the test

192

Kanematsu et al.

192 Kanematsu et al. Figure 3. A 64-year-old man with diffuse-type HCC in hemo- chromatosis, showing

Figure 3. A 64-year-old man with diffuse-type HCC in hemo- chromatosis, showing negative serum alpha-fetoprotein value ( 10 ng/mL). A: Unenhanced T2-weighted short tau inversion recovery turbo spin-echo (5,110/76/170 [TR/TE/TI]) axial MR image shows the tumor involving most of the left hepatic lobe and a part of right hepatic lobe as areas of homogeneous, moderate hyperintensity (arrow). Note extensive portal venous tumor thrombosis (curved arrow). B: Gadolinium-enhanced early-phase spoiled gradient-echo (160/4.5) axial MR image shows the tumor as areas of intense, patchy enhancement (arrow). Note the very low signal intensity of the liver paren- chyma (curved arrow), reflecting underlying hemochromato- sis.

result was available (0.7–23.5 mg/dL [mean, 6.5 8.1 mg/dL]), serum AFP value (0 –10 ng/mL) was elevated in 14 of 18 patients (5–242000 ng/mL, mean; 24822 56380 ng/mL), and 13 showed a value greater than 500 ng/mL. Serum alkaline phosphatase value (38 –126 U/liter) was elevated in 17 of 20 patients (99 –969 U/li- ter [356 267 U/liter]), serum alanine aminotransfer- ase (15–48 U/liter) value was elevated in 14 of 20 pa- tients (21–190 U/liter [84 52 U/liter]), serum aspartate aminotransferase value (14 –38 U/liter) was elevated in 18 of 20 patients (38 –876 U/liter [243 226 U/liter]), serum -glutamyltranspeptidase value (11–48 U/liter) was elevated in all of 18 patients (54 – 1042 U/liter [361 270 U/liter]). Number of Couinaud’s segments involved by diffuse- type HCC was four segments in four patients, five seg- ments in one, six segments in four, seven segments in

one, and eight segments in 12 (mean, 6.7 1.6 seg- ments). Fibrous capsules or septa were seen mildly but partly in seven, minimally in two, and none in 13. Unenhanced T1-weighted MR images showed the tu- mors as areas of homogeneous hypointensity in five patients, heterogeneous hypointensity in ten, mixed in- tensity in three, and isointensity in four. T2-weighted MR images showed the tumors as areas of heteroge- neous hyperintensity in 16 and homogeneous hyperin- tensity in six. Postcontrast early-phase MR images showed tumors as areas of patchy enhancement in 12 patients, miliary enhancement in nine, and minimal enhancement in one. Postcontrast late-phase MR im- ages showed heterogeneous irregular areas of wash-out in all patients. Portal venous tumor thrombosis was seen in all pa- tients: bilateral main portal vein branches were in- volved in 12 patients and either of the right or left main portal vein branch was involved in 10. Hepatic venous tumor thrombosis was not seen in any patient. Intra- hepatic bile ducts were minimally to mildly dilated in three patients and of normal caliber in 19. Ascites was substantial in one patient, moderate in seven, minimal in eight, and absent in six. Upper ab- dominal lymphadenopathy was seen in three patients (porta-hepatis nodes in all three and one with addi- tional porto-caval and peripancreatic nodes). Distant metastases were confirmed in three patients (bone in one, adrenal gland in one, and lung in one).

DISCUSSION

A 1981 report by Okuda et al (8) described the clinical and pathological findings with six autopsy cases of dif- fuse-type HCC. In the report, the prominent clinical feature was the rapid deterioration of the patient’s gen- eral condition terminating in hepatic failure, and the liver size enlarged quickly at a perceptible speed, often accompanied by abdominal pain. They also described that the entire liver was studded with minute, uni- formly sized tumor nodules evenly distributed through- out, with some of them grossly indistinguishable from cirrhotic nodules. The incidence of diffuse-type HCC in our institution was approximately 13% of all patients with HCC, which is slightly higher than previously reported (8). This may reflect differences related to geography, as our study involved a North American population, variation related to the relative rarity of the disease, or that our study was imaging-based rather than autopsy-based. Portal venous tumor thrombosis is a common finding with diffuse-type HCC and may be a clue to the diag- nosis. Accordingly, characterization of portal venous thrombosis is crucial in the diagnosis of diffuse-type HCC on contrast-enhanced CT or MR imaging. Tublin et al (9) evaluated 58 cirrhotic patients and concluded that identification of main portal venous thrombosis of 23 mm in diameter or greater or neovascularity of portal venous thrombosis resulted in a sensitivity and speci- ficity for the CT characterization of malignant portal venous thrombosis of 86% and 100%, respectively. No previous reports have described the specific MR imaging findings of diffuse-type HCC cases. We found that the unenhanced T1- and T2-weighted MR images

MRI of Diffuse HCC

Table 1 Summary of Patient Characteristics in 22 Patients With Diffuse HCC

193

Patients

Age/sex/race

Etiology of underlying liver disease

Cirrhosis

TB

(mg/dl)

AFP

(ng/mL)

ALP

(U/liter)

ALT

(U/liter)

AST

(U/liter)

-GTP

(U/liter)

No. of

involved

segments

Capsules

or septa

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

60/M/C

30/M/AA

51/M/C

48/M/AA

46/M/C

50/M/AA

64/M/C

49/M/A

49/M/AA

48/M/C

49/M/AA

51/M/AA

49/M/C

44/M/C

61/F/C

43/M/C

58/M/AA

64/M/C

16/F/C

64/M/C

71/M/C

80/F/C

Alcohol abuse

Alcohol abuse

Alcohol abuse

Alcohol abuse

Alcohol abuse

Hep. B

Hep. B

Hep. B

Hep. C

Hep. C

Hep. C

Hep. C

Hep. C

Hep. C

Hep. C

Alcohol abuse, Hep. C

Hep. B, Hep. C

Hemochromatosis

Autoimmune hepatitis

Cryptogenic

Cryptogenic

Cryptogenic

Yes

Yes

Yes

Yes

Yes

No a

Yes

No

a

No

a

Yes

Yes

Yes

No a

Yes

Yes

Yes

Yes

No a

Yes

Yes

No a

Yes

3.1

23.5

19.3

1.3

1.5

20.8

3.3

N.A.

8.7

2.2

1.8

2.7

2.3

2.0

2.0

2.2

5.2

0.7

N.A.

23.5

2.1

1.9

38,000

15,870

38,000

2,000

N.A.

564

5366

N.A.

42,000

1816

N.A.

136

5

9000

242,000

7000

38,000

10

N.A.

7000

10

10

242

114

99

245

551

623

851

N.A.

165

310

173

195

122

168

784

174

280

311

N.A.

193

969

548

136

28

53

66

28

155

147

N.A.

47

21

40

132

134

50

60

53

113

55

N.A.

190

134

46

395

65

118

105

38

700

404

N.A.

207

67

38

447

162

268

257

94

292

73

N.A.

876

168

86

257

350

84

274

666

177

1042

N.A.

184

445

139

135

254

N.A.

N.A.

249

288

630

N.A.

54

822

451

4 segs.

4 segs.

All

All

All

All

All

6 segs.

All

5 segs.

All

7 segs.

4 segs.

All

All

6 segs.

4 segs.

6 segs.

All

All

6 segs.

All

a Patients had chronic hepatitis. TB serum total bilirubin (normal range, 0–1.2 mg/dl), AFP serum alpha-fetoprotein (0–10 ng/mL), ALP serum alkaline phosphatase (38–126 U/liter), ALT serum alanine aminotransferase (15–48 U/liter), AST serum aspartate aminotransferase (14–38 U/liter), -GTP serum -glutamyltranspeptidase (11–48 U/liter), C Caucasian, AA African-American, A Asian, Hep. B type-B viral hepatitis, Hep. C type-C viral hepatitis, N.A. not available, mild, minimal, – none.

Table 2 Summary of MRI Findings in 22 Patients With Diffuse HCC

Patients

T1-weighted imaging

findings

T2-weighted imaging

findings

Enhancement pattern

in early-phase

postcontrast images

Washed-out area % in late-phase postcontrast images

Portal venous

tumor thrombosis

in MPV(s)

Bile duct

dilatation

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

Iso

Iso

Homo, mild hypo

Hetero, mild hypo

Hetero, mild hypo

Hetero, mild hypo

Hetero, mod. hypo

Hetero, mild hypo

Homo, mod. hypo

Mixed

Hetero, mild hypo

Hetero, mild hypo

Homo, mild hypo

Homo, mild hypo

Hetero, mild hypo

Hetero, mod. hypo

Iso

Hetero, mod. hypo

Iso

Mixed

Mixed

Homo, mod. hypo

Hetero, mild hyper

Hetero, mild hyper

Hetero, mod. hyper

Hetero, mild hyper

Hetero, mod. hyper

Hetero, mild hyper

Hetero, mild hyper

Hetero, mild hyper

Homo, mod. hyper

Homo, mild hyper

Hetero, mild hyper

Hetero, mild hyper

Homo, mild hyper

Hetero, mod. hyper

Hetero, mild hyper

Hetero, mild hyper

Homo, mod. hyper

Homo, mod. hyper

Homo, mod. hyper

Hetero, mild hyper

Hetero, mild hyper

Hetero, mod. hyper

Miliary

Patchy

Patchy

Miliary

Patchy

Miliary

Patchy

Miliary

Miliary

Patchy

Patchy

Patchy

Patchy

Miliary

Miliary

Patchy

Minimal

Patchy

Patchy

Patchy

Miliary

Miliary

50–75%

50–75%

75–100%

25–50%

75–100%

75–100%

50–75%

75–100%

75–100%

25–50%

50–75%

25–50%

25–50%

25–50%

50–75%

75–100%

50–75%

25–50%

50–75%

25–50%

50–75%

50–75%

Right

Right

Bilateral

Left

Bilateral

Bilateral

Bilateral

Right

Bilateral

Right

Right

Bilateral

Right

Bilateral

Bilateral

Right

Left

Bilateral

Bilateral

Bilateral

Right

Bilateral

MPV(s) main portal vein branch(es), mild, minimal, – none, Homo homogeneous, Hetero heterogeneous, mod. moderate, hypo hypointensity, hyper hyper intensity, iso iso intensity.

194

Kanematsu et al.

showed the tumors as ill-demarcated areas of homoge- neous or heterogeneous, abnormal signal intensities, typically mildly to moderately hypointense on T1- weighted images and mildly to moderately hyperintense on T2-weighted images. We considered the findings on noncontrast images somewhat nonspecific as these sig- nal alterations may also be analogous to those in cir- rhotic liver without tumor. However, multiple small ar- eas of low signal intensity in the liver on T2-weighted images may suggest the presence of regenerating nod- ules in cirrhosis, which may help exclude the diagnosis of diffuse-type HCC (10). We observed patchy or miliary enhancement corre- sponding to the tumors on the early-phase postcontrast images. Miliary enhancement in particular, we believe, may be relatively specific for diffuse-type HCC, and may represent the enhancement of extensive micronodules of tumor as identified at histopathology (8). Heteroge- neous wash-out of the tumors on the late-phase images is a feature typical for malignant tumor of all types in the liver. Increased hepatic parenchymal enhancement on early-phase postcontrast images is known to be caused by increased hepatic arterial supply in the set- ting of proximal portal vein obstruction (11), but this usually is observed as homogeneous, wedge-shaped enhancement in postcontrast early phase, with fading to homogeneous isointensity in postcontrast late phase. We could distinguish such non-neoplastic, tran- sient early-phase enhancement, which was seen in some cases, from the enhancement of tumor. In our clinical experience diffuse HCC may be distin- guished from other malignant diseases, specifically cholangiocarcinoma and metastases, in the great ma- jority of cases. Cholangiocarcinoma, in our experience, has better defined margins and we have not observed portal vein tumor thrombus with this entity, although portal vein comparison is common. Massive involve- ment of the liver in metastatic disease almost invariably shows a more clearly focal pattern of involvement. In our experience, breast cancer is the malignancy that most often may present with extensive infiltration. In this setting the primary tumor is almost always known. Coexistence of breast cancer and cirrhosis is rare but may occur, so uncertainty whether the patient has breast cancer liver metastases or diffuse HCC is rare. Finally, while venous tumor thrombus was seen in all patients with diffuse HCC, it is rare in patients with metastatic disease. Lee et al (12) reported that the serum AFP level, which showed a positive predictive value of 95% for HCC, was 3200 ng/mL in livers with type-B viral hepatitis and 200 ng/mL in livers without type-B viral hepatitis. They concluded that the presence of underlying type-B viral hepatitis should be taken into consideration when se- rum AFP is measured to diagnose HCC. One of our patients had underlying type-B viral hepatitis and had a serum AFP value of 5366 ng/mL, which was consid- ered diagnostic for HCC. Serum AFP values have been reported to be elevated in 43%–72% of patients with HCC (13–16). However, as more small HCCs are currently being detected, due to the advent of advanced radiologic imaging techniques

and increased imaging screening of patients with hep- atitis or cirrhosis, this rate is likely much lower, as small well-differentiated focal HCCs rarely result in el- evated AFP (17). We observed that 14 (78%) of 18 pa- tients with diffuse HCC showed elevated serum AFP values in our study. Radiologists should be apprised of the fact that although the positive rate of AFP is high with diffuse HCC, a sizable number (22% in our study) have normal serum AFP value. There are some limitations to this study. We did not have a histopathologic diagnosis of HCC in eight pa- tients. These patients were very ill and the combination of the MR findings and markedly elevated serum AFP values ( 2000 ng/mL) were considered diagnostic for HCC, as also described in the literature (12). In conclusion, diffuse-type HCC represented 13% of the patients with HCC. Diffuse-type HCC was seen as extensive permeative hepatic tumor with ill-defined margins, with extensive portal venous tumor thrombo- sis in all patients, and markedly elevated serum AFP value in 78% of patients. Serum AFP value was normal in 22% of patients. Unenhanced MR images showed nonspecific findings of abnormal signal intensities. Ga- dolinium-enhanced MR images showed patchy or mili- ary enhancement on immediate postcontrast images, with the miliary pattern possessing a distinctive ap- pearance.

REFERENCES

  • 1. Tsukuma H, Hiyama T, Tanaka S, et al. Risk factors for hepatocel- lular carcinoma among patients with chronic liver disease. N Engl J Med 1993;328:1797–1801.

  • 2. National Institutes of Health Consensus Development Conference Panel statement: Management of hepatitis C. Hepatology 1997;26: 2–10.

  • 3. Donato F, Tagger A, Chiesa R, et al. Hepatitis B and C virus infec- tion, alcohol drinking, and hepatocellular carcinoma: a case-con- trol study in Italy. Brescia HCC study. Hepatology 1997;26:579 –
    584.

  • 4. Niederau C, Fischer R, Sonnenberg A, Stremmel W, Trampisch HJ, Strohmeyer G. Survival and causes of death in cirrhotic and non- cirrhotic patients with primary hemochromatosis. N Engl J Med 1985;313:1256 –1262.

  • 5. Bradbear RA, Bain C, Siskind V, et al. Cohort study of internal malignancy in genetic hemochromatosis and other non-alcoholic liver disease. J Natl Cancer Inst 1985;75:81–84.

  • 6. Okuda K, Peters RL, Simson IW. Gross anatomic features of hepa- tocellular carcinoma from three disparate geographic areas. Pro- posal of new classification. Cancer 1984;54:2165–2173.

  • 7. Anthony PP. Tumours and tumour-like lesions of the liver and biliary tract: aetiology, epidemiology, and pathology. In: MacSween RNM, Burt AD, Portmann BC, et al., eds. Pathology of the liver. 4th ed. London, England: Churchill Livingstone; 2002. p 711–825.

  • 8. Okuda K, Noguchi T, Kubo Y, Shimokawa Y, Kojiro M, Nakashima T. A clinical and pathological study of diffuse type hepatocellular carcinoma. Liver 1981;1:280 –289.

  • 9. Tublin ME, Dodd 3rd GD, Baron RL. Benign and malignant portal vein thrombosis: differentiation by CT characteristics. AJR Am J Roentgenol 1997;168:719 –723.

  • 10. Itai Y, Ohnishi S, Ohtomo K, et al. Regenerating nodules of liver cirrhosis: MR imaging. Radiology 1987;165:419 –423.

  • 11. Itai Y, Moss AA, Goldberg HI. Transient hepatic attenuation differ- ence of lobar or segmental distribution detected by dynamic com- puted tomography. Radiology 1982;144:835–839.

  • 12. Lee HS, Chung YH, Kim CY. Specificities of serum alpha-fetopro- tein in HBsAg and HBsAg- patients in the diagnosis of hepatocel- lular carcinoma. Hepatology 1991;14:68 –72.

MRI of Diffuse HCC

195

13. Sherman M, Peltekian KM, Lee C. Screening for hepatocellular carcinoma in chronic carriers of hepatitis B virus: incidence and

prevalence of hepatocellular carcinoma in a North American urban population. Hepatology 1995;22:432–438. 14. Aoyagi Y, Oguro M, Yanagi M, et al. Clinical significance of simultaneous determinations of alpha-fetoprotein and des- gamma-carboxy prothrombin in monitoring recurrence in pa- tients with hepatocellular carcinoma. Cancer 1996;77:1781–

1786.

15. Okuda H, Nakanishi T, Takatsu K, et al. Measurement of serum

levels of des-gamma-carboxy prothrombin in patients with hepato-

cellular carcinoma by a revised enzyme immunoassay kit with increased sensitivity. Cancer 1999;85:812–818.

  • 16. Nguyen MH, Garcia RT, Simpson PW, Wright TL, Keeffe EB. Racial differences in effectiveness of alpha-fetoprotein for diagnosis of hepatocellular carcinoma in hepatitis C virus cirrhosis. Hepatology 2002;36:410 –417.

  • 17. Peterson MS, Baron RL, Marsh Jr JW, Oliver 3rd JH, Confer SR, Hunt LE. Pretransplantation surveillance for possible hepatocellu- lar carcinoma in patients with cirrhosis: epidemiology and CT- based tumor detection rate in 430 cases with surgical pathologic correlation. Radiology 2000;217:743–749.